Skip to main content

Full text of "Greek Philosophy Part I Thales To Plato"

See other formats

180 B96 60-07229 
Greek philosophy 

* H ^ 

' 6 

of flijiiosopijp 

















First Edition 1914. 
Reprinted 1920, 1924. 



THE preparation of this volume was undertaken some 
years ago, but was interrupted by my work on the Lexicon 
Platonicum, which has proved a more formidable task than 
was at first anticipated. I have to thank the editor of this 
series and the publishers for their generous indulgence in 
the circumstances. 

It is unfortunate in some respects that I have been 
obliged to deal with certain parts of the subject in a form 
which does not admit of detailed argument and still less 
of controversy. The second edition of my Early Greek 
Philosophy (referred to as E. Gr. Phf} makes this in large 
measure unnecessary in Book L, but there are certain parts 
of Book III. where I have had to state my conclusions 
baldly in the hope that I may have a later opportunity 
of discussing their grounds. My chief aim for the present 
has been to assist students who wish to acquire a firsthand 
knowledge of what Plato actually says in the dialogues of 
Jiis maturity. So long as they are content to know some- 
thing of the Republic and the earlier dialogues, Platonism 
must be a sealed book to them. 

I have not thought it well to present Greek names in a 
Latin dress. I see no advantage, and many disadvantages, 
in writing Herakleitos as Heraclitus. It often leads to his 
being called out of his name, as the Emperor Herakleios 


usually is when disguised as Heracttus. On the other 
hand, the Latin titles of Plato's dialogues are English 
words. Theaitetos of Athens is best left with the beautiful 
name chosen for him by his father Euphronios, but cc the " 
Theaetetus is as much English as Thessalonians. We shall 
f never, it seems, reach agreement on this matter ; I only 
wish to explain my own practice. 

I have to thank my friend and former colleague, Sir 
Henry Jones, for many valuable suggestions and, above 
all, for his constant encouragement. Mr. Hetherington 
of Glasgow University was good enough to verify most 
of my references, and the proofs have been carefully read 
by Mr. W. L. Lorimer, Lecturer in Greek at the Univer- 
sity of St. Andrews. For the imperfections which remain 
I am solely responsible. 

J. B. 



INTRODUCTION- ------- i 



THE IONIANS -------- 17 

Miletos .-- 17 

The Breakdown of Ionian Civilisation 28 

Religion 3 1 

Enlightenment .------32 

PYTHAGORAS --------37 

The Problem - 37 

Life and Doctrine ------ 38 

Music --------45 

Medicine 49 

Numbers - - - - - - - - 5 1 



Herakleitos --------57 

Parmenides -------- 63 



THE PLURALISTS ._--_._ 69 

Empedokles- - - - - - - - 71 

Anaxagoras -------- 76 



Zeno - ....... -82 

Melissos ------- -85 

The Later Pythagoreans - - - - - 87 

LEUKIPPOS .-.,--... .. 



THE SOPHISTS ....... . 105 

Law and Nature- -----. 105 

The "Sophists" ....... 107 

Protagoras - - - ..... no 

Hippias and Prodikos - - - - - - 118 

Gorgias - - - - - - - -119 

Eclectics and Reactionaries - - - - -122 


The Problem - - - - - - -126 

The Platonic Sokrates - - - - - -128 

Aristophanes and Xenophon - - - - - 144 





The Associates of Sokrates - - - - -151 

The Forms - - - - - * - 154 

Goodness - - - - - ~ - -170 


The Condemnation - - - - - -180 

The Alleged Offence - - - - - 182 

The Real Offence 185 

The Pretext 189 

The Death of Sokrates - - - - -191 


DEMOKRITOS _---_.._ 193 
Theory of Knowledge- - - - - 196 

Theory of Conduct - - - - - -199 




Plato's Early Life --.... 205 

Foundation of the Academy - - - - 213 

Plato and Isokrates - - - - - -215 

The Methods of the Academy - - - 219 

The Programme of Studies ----- 223 

Eukleides and Plato ---,.-. 230 



CRITICISM .---_-.. 234 

The Theaetetus - - - - - - -237 

The Parmenides - - - - - - ~ 2 53 


LOGIC -_.-.---- 273 

The Sophist ------- 273 


POLITICS --------- 290 

The Statesman ------- 290 

Plato and Dionysios ------ 294 

The Laws -------- 301 

Education -------- 305 



I. Forms, Mathematicals and Sensibles - - 315 

II. The One and the Indeterminate Dyad - - 320 

The Philebus 324 



The Soul ..------ ~ 333 

God --------- 335 

The World ------- 338 

Conclusion ------- - 349 


INDEX --...-.-. 352 


Jo one will ever succeed in writing a history of philo- 
ophy ; for philosophies, like works of art, are intensely 
>ersonal things. It was Plato's belief, indeed, that no 
>hilosophical truth could be communicated in writing at 
11 ; it was only by some sort of immediate contact x that 
me soul could kindle the flame in another. Now in dealing 
vith the philosophy of an earlier age, we are wholly con- 
ined to written records, and these are usually fragmentary 
ind often second-hand or of doubtful 'authority. They 
ire written, too, in a language which at best we only half 
anderstand, and have been moulded by influences for the 
most part beyond our ken. It will only, therefore, be in 
so far as the historian can reproduce the Platonic contact 
of souls that his work will have value. In some measure 
this is possible. Religious faith often seems able to break 
through the barriers of space and time, and so to appre- 
hend its object directly ; but such faith is something 
personal and incommunicable, and in the same way the 
historian's reconstruction of the past is primarily valid for 
himself alone. It is not a thing he can hand over ready- 
made to others. There is nothing mysterious about this 
aspect either of religious faith or of philological inter- 
pretation. On the contrary, all knowledge has the same 
character. In the present case it only means that a man who 
tries to spend his life in sympathy with the ancient philo- 
sophers 1 will sometimes find a direct conviction forcing itself 

1 This is what Plato calls rb crvffiv (Ef. vii. 34.1 c), but he is thinking 
of the living, not the dead. 


upon him, the grounds of which can only be represented 
very imperfectly by a number of references in a footnote. 
Unless the enumeration of passages is complete and it 
can never be complete and unless each passage tells 
exactly in the same way, which depends on its being read 
in the light of innumerable other passages not consciously 
present to memory, the so-called proofs will not produce 
the same effect on any two minds. That is the sense 
in which philological inquiry, like every other inquiry, 
requires an act of faith. It is clear, however, that no one 
whose experience has not been identical can be called 
on to repeat this act after another, and for this reason 
professed histories of philosophy are often more of a 
hindrance than a help. They seem only to interpose 
f another obstacle where there are obstacles enough already. 
But though a history of philosophy is impossible, there 
are some humbler tasks that can in a measure be per- 
formed, and of which the performance may help to 
prepare the way for a more direct vision. In the first 
place, there are certain external matters that may be 
determined with considerable accuracy and which are not 
without importance. We are more likely to understand a 
philosopher rightly if we know the time he lived at and 
the surroundings that may have helped to shape his 
thought, even though these can never wholly explain him. 
It is particularly useful to know what other philosophers 
he was acquainted with, either directly or through their 
writings. In the second place, the development of Greek 
philosophy depends on the progress of scientific, and 
especially mathematical, discovery more than on anything 
else, and it is possible to ascertain pretty accurately the 
stage Greek science had reached by a given time. The 
records are full, and, when critically used, trustworthy. 
It is for these reasons that this work deals so largely with 
matters which may appear at first to lie outside the pro- 
vince of philosophy. That is, in fact, its chief justification. 
It is an attempt to lead the reader to the right point of 
view, from which he may then see for himself. Lastly, 


there is what may be called the cathartic or purgative 
function of history. The greatest of all the obstacles we 
have to surmount is just the mass of scholastic explana- 
tion and dogma which so soon overwhelm the teaching of 
any original genius. To clear that away is perhaps the 
greatest service that can be rendered in this field. We 
do not wish to see Plato with the eyes of Aristotle, 
or even of Plotinos, but if possible, face to face, and 
anyone who can help us here deserves our thanks. It 
may seem a purely negative service, but that lies in the 
nature of the case. In the long run the positive con- 
struction must be left to the individual student, and no 
two students will see quite alike. All the historian can 
do is to point the way, and warn others off tracks which 
have already been found to lead nowhere. 

Even this, however, implies that we know already what 
philosophy is, and clearly, unless we have some notion of 
that, we shall be in danger of losing the thread of our 
story. We can nevertheless dispense with such a defini- 
tion as would be applicable to the philosophy of all ages 
and peoples, for we shall find a pretty clear notion of 
what philosophy was during the Hellenic period emerging 
as we go on. This will at least do justice to one aspect 
of the subject, and that the one we are immediately con- 
cerned with. It will be convenient to state at once, 
however, that for the purpose of this work, I mean by 
philosophy all Plato meant by it, and nothing he did not 
mean by it. The latter point is important ; for it means 
that philosophy is not mythology, and, on the other hand, 
that it is not positive science, however closely it may be 
related to both of these. 


In the first place, philosophy is not mythology. It is 
true that there is plenty of mythology in Plato, and we 
shall have to consider the meaning of that later. It is 
also true that we shall have to take account from the first 


of a mass of cosmogonical and eschatological speculation 
which influenced philosophy in many ways. These things, 
however, are not themselves philosophy, and it cannot 
even be said that they are the germ from which philosophy 
developed. It is important to be quite clear about this ; 
for in some quarters Oriental cosmogonies are still paraded 
as the source of Greek philosophy. The question is not 
one of cosmogonies at all. The Greeks themselves had 
cosmogonies long before the days of Thales, and the 
Egyptians and Babylonians had cosmogonies that may be 
older still. Even savages have cosmogonies, and they are 
nearly as advanced as those of more civilised peoples. It 
is possible, though it has certainly not been proved, that 
the oldest Greek cosmogonies, or some of them, came from 
Egypt or Babylon. It is still more probable that systems 
such as that of Pherekydes have preserved fragments of 
"Minoan" speculation, which may be of indefinite 
antiquity. These things, however, have nothing directly 
to do with philosophy. From the Platonic point of view, 
there can be no philosophy where there is no rational 
science. It is true that not much is required a few pro- 
positions of elementary geometry will do to begin with 
but rational science of some sort there must be. Now 
rational science is the creation of the Greeks, and we know 
when it began. We do not count as philosophy anything 
anterior to that. 


It is true, of course, that science originated at the time 
when communication with Egypt and Babylon was easiest, 
and just where the influence of these countries was likely 
to be felt, and it is a perfectly fair inference that this had 
something to do with its rise. On the other hand, the 
very fact that for two or three generations Greek science 
remained in some respects at a very primitive stage affords 
the strongest presumption that what came to Hellas from 
Egypt and Babylon was not really rational science. If the 


Egyptians had possessed anything that could rightly be 
called mathematics, it is hard to understand how it was 
left for Pythagoras and his followers to establish the most 
elementary propositions in plane geometry ; and, if the 
Babylonians had really any conception oif the planetary 
system, it is not easy to see why the Greeks had to dis- 
cover bit by bit the true shape of the earth and the ex- 
planation of eclipses. It is clear that these things were 
not known at Babylon ; they were gradually worked out 
in South Italy, where we can hardly assume Oriental 
influences. Of course everything depends on what we 
mean by science. If we are prepared to give that name 
to an elaborate record of celestial phenomena made for 
purposes of divination, then the Babylonians had science 
and the Greeks borrowed it from them. Or, if we are 
prepared to call rough rules of thumb for measuring 
fields and pyramids science, then the Egyptians had 
science, and it came from them to Ionia. But, if we 
mean by science what Copernicus and Galileo and Kepler, 
and Leibniz and Newton meant, there is not the slightest 
trace of that in Egypt or even in Babylon, while the very 
earliest Greek ventures are unmistakably its forerunners. 
Modern science begins just where Greek science left off, 
and its development is clearly to be traced from Thales to 
the present day. Copernicus says himself that he was 
put on the track by what he read of the Pythagoreans in 
the Placita ascribed to Plutarch, 1 

The only remains that have come down to us show that 
the Egyptians were not without a certain ingenuity in 
the solution of particular arithmetical and geometrical 
problems, but there is not the slightest trace of anything 
like general methods. 2 If inconvenient remainders occur, 
they are simply dropped. In the same way, the rules 

1 j5. Gr. Ph? p. 349, n. 2. It was " the Pythagorean doctrine, taught 
also by Nicolas Copernicus," that was condemned by the Congregation 
of the Index in 1616. 

2 For the Rhind papyrus, see E. Gr. Ph. 2 pp. zzfF,, and, for a later 
discussion, see v. Bissing in Neue Jahrbttcher, xxv. (1912), pp. 81 ff. 


given for reducing triangles to rectangles are only correct 
if the triangles are right-angled, though those given in the 
diagrams are apparently meant to be equilateral. In fact 
the whole system resembles the rough and ready methods 
of the Roman agrimensores far more than anything we 
should call scientific. Nor is there the slightest ground 
for the statement sometimes made that the Egyptians had 
a more highly developed geometry which they guarded as a 
mystery. That is based mainly on the story that Plato 
went to Memphis to study under the priests, a story 
for which there is no good evidence. In any case we 
know Plato's opinion of Egyptian mathematics, and it is 
that there was an element of illiberality in it due to its 
preoccupation with merely practical ends. 1 It is stated 
that, though hexagons are common on the Egyptian 
monuments, the pentagon is never found. 2 If that is so, 
it is very significant. Anyone can make hexagons, but 
the construction of the regular pentagon is a different 
matter. We shall see that it was known to the Pytha- 
goreans, to whom the pentagon was of interest as the side 
of the regular dodecahedron, the most important figure 
in their system. It should be added that all mathematical 
terms, * pyramid ' included, are of pure Greek origin. 8 

It is true, of course, that in Hellenistic times, a certain 
number of Egyptian priests applied the methods of Greek 
science to the traditional lore of their own country. The 
Hermetic literature proves it, and so does the elaborate 
astrological system the later Egyptians erected on a Stoic 
foundation. All that, however, throws no light on the 
origins of Greek science. On the contrary, if the Egyptians 
of these days adopted the contemporary Greek science 

1 Plato, Laws, 747 b, 6 sq q . 

2 Zeuthen, Histoire des mathtmatlques (Paris 1902), p. 5. 

8 The words irvpafus, Trvpa/zofis, which mean a cake made of wheat 
and honey, are clearly derived from Trvpoi, * wheat/ though their form 
has been influenced by the analogy of cnjo-a/us, onycra^ous. See also 
E. Gr. Ph? p. 25, n. i. 


and philosophy, it is only another indication of their own 
poverty in such things. 


In the case of Babylon it is even more important to 
distinguish the times before and after Alexander the 
Great. In the latter period Babylon had become a 
Hellenistic city, and there was free intercourse between 
the astronomers of Mesopotamia and Alexandria. It is 
certain that Hipparchos, for instance, made use of Baby- 
lonian observations. But Greek science was fully consti- 
tuted before his time, and there can hardly be any doubt 
that Babylonian astronomy attained its highest develop- 
ment under Greek influence. 1 What we have really to 
consider is whether there is any trace of it in Hellas at a 
much earlier date. Now we know a few facts about this, 
and they are instructive. According to Herodotos (ii. 
109), it was from Babylon the Greeks got the instru- 
ment called the gnomon^ which indicated the solstices 
and equinoxes by a shadow. Whether that is a scientific 
instrument or not depends on what you do with it. 
The Greeks were also familiar at an early date with the 
Babylonian duodecimal and sexagesimal systems of 
numeration, but the use of these was limited to weights, 
measures, and currency, or, in other words, to com- 
mercial purposes. They were not employed in science 
till Hellenistic times, when the circle was divided into 
degrees. Arithmetic proper used only the decimal 
system. If they had cared, the Greeks might have 
learned from the Babylonians to distinguish the planets. 
These were of the greatest importance for purposes of 
divination, but the Greeks paid no attention to astrology 
before the third century B.C. 2 So long as there was no 

1 For recent statements on this subject, see Jastrow in Enc. Brit, (nth 
edition), vol. ii. pp. 796 f. ; Boll in Neue Jahrbticher, xxi. (1908), p. 116. 

2 See Cumont in Neue Jahrbilcher, xxiv. (191 1), pp. i ff. He says (p. 4) : 
"The universal curiosity of the Hellenes by no means ignored astrology, 


cosmological system in which the "tramp-stars 
as the Greeks irreverently called them, could find a place, 
they did not strike them as of more consequence than 
shooting stars and the like. The Pythagoreans appear to 
have worked out their planetary theory quite indepen- 
dently after discovering the real nature of the earth. It 
was said to be Pythagoras or Parmenides that first 
identified the evening and the morning star. The Greek 
equivalents for the Babylonian names of the planets, which 
we still use in their Latin form, appear for the first time 
in the Platonic Epinomis (987 b sq!). Evidently, then, the 
Greeks did not learn from the Babylonians the single piece 
of real astronomical knowledge they possessed. 

They did, however, make use of one important achieve- 
ment of theirs in this field, namely, their records of 
eclipses, and the various cycles established on the basis of 
these records. They used these for the purposes of the 
calendar, and, as we shall see, for the prediction of 
eclipses. Whether such observations and calculations are 
scientific or not depends wholly on the purpose with 
which they are made and the uses to which they are put. 
In itself an eclipse of the sun is a phenomenon of purely 
local interest, and it is no more scientific to record it than 
it would be to record rainbows. If the record suggests 
that something has really happened to the sun, and that 
something may therefore happen to the King, it is not 
only not science, but an instrument of positive nescience. 
That, however, was the view taken by the astronomers of 

The only eastern people that can bear comparison with 
the Greeks in science and philosophy are the Indians. 

but their sober understanding rejected its adventurous doctrines. Their 
acute critical sense knew well how to distinguish between the scientific 
observations of the Chaldeans and their erroneous inferences. It remains 
their everlasting glory that they discovered and made use of the serious, 
scientific elements in the confused and complex mass of exact observa- 
tions and superstitious ideas, which constitutes the priestly wisdom of the 
East, and threw all the fantastic rubbish on one side/'" 


How much of Indian science is original, and how much 
may be traced to Greek influence, is a very difficult ques- 
tion in view of the uncertainty of Indian chronology. It 
does seem certain, however, that no Indian scientific work, 
and therefore nothing we count as philosophy, can be 
dated with probability before the time of Alexander. In 
particular, there is no ground for believing that the mathe- 
matical book entitled the Suha-sutras, or "rules of the 
cord," is of earlier date, and it is in any case far below the 
level of Greek science. 1 The analogy of Egypt and 
Babylon certainly suggests that this reached India from 
the Hellenistic kingdom of the North West 

The truth is that we are far more likely to underrate 
the originality of the Greeks than to exaggerate it, and we 
do not always remember the very short time they took to 
lay down the lines scientific inquiry has followed ever 
since. By the early part of the sixth century B.C. they 
had learnt the rough and ready system of mensuration 
which was all Egypt could teach them, and a hundred 
years later we find the study of arithmetical and geo- 
metrical progressions, plane geometry and the elements of 
harmonics firmly established on a scientific basis. Another 
century saw the rise of solid and spherical geometry, and 
the sections of the cone were soon added. The Greeks 
learnt, directly or indirectly, from Babylon that certain 
celestial phenomena recur in cycles, and may therefore be 
predicted. Within fifty years they had discovered that 
the earth swings free in space, and the knowledge of its 
spherical shape soon followed. A century saw the true 
account of eclipses clearly stated, and this led up to the 

1 See A. B. Keith in the Journal of the Royal Asiatic Society, 1909, 
pp, 589fF. It is a pity that M. Milhaud has been persuaded to accept 
an early date for the Sulva-sutras in his Nouvelks ttudes (1911), pp. 


discovery that the earth was a planet. A little later some 
Greeks even taught that the sun was not a planet, but the 
centre of the planetary system. Nor must we forget that 
hand in hand with this remarkable development of mathe- 
matical and astronomical science there went an equally 
striking advance in the study of the living organism. 
Most of the writings that have come down to us under 
the name of Hippokrates belong to the fifth century B.C., 
and, while some of them show a tendency to the specula- 
tive interpretation of vital phenomena natural in an age of 
rapid scientific progress, there are others which display in 
an almost perfect form the method of minute and pains- 
taking observation that is alone appropriate in dealing 
witfc facts of such complexity. The physicians of Alex- 
andria discovered the nervous system, but the native 
Egyptians, though accustomed for some thousands of 
years to embalm dead bodies, show astounding ignorance 
of the simplest anatomical facts. 

The Greeks achieved what they did, in the first place, 
because they were born observers. The anatomical 
accuracy of their sculpture in its best period proves that, 
though they never say anything about it in their literature, 
apparently taking it for granted. The Egyptians, we 
may remember, never learnt to draw an eye in profile. 
But the Greeks did not rest content with mere observa- 
tion ; they went on to make experiments of a quite 
modern character. That by which Empedokles illustrated 
the flux and reflux of the blood between the heart and the 
surface of the body is the best known ; for we have a 
description of it in his own words. 1 It also established 
the corporeal nature of atmospheric air. We should 
certainly hear of many more such experiments if our 
sources were less meagre, and more intelligently compiled. 
Further, the Greeks always tried to give a rational 
explanation (\oyov SiSovai) of the appearances they had 
observed. Their reasoning powers were exceptional, as 
we can see from the mathematical work they have left us. 

. Gr. Ph? p. 253. 


On the other hand, they were also quite conscious of the 
need for verification. This they expressed by saying 
that every hypothesis must "save the appearances" 
(o-f^eiv ra (paivojmeva) ; in other words, that it must do 
justice to all the observed facts. 1 That is the method 
of science, as we understand it still. It should be added 
that the development of mathematical and biological 
science at a given time to a large extent determines 
the character of its philosophy. We shall see how the 
mathematical influence culminates in Plato, and the bio- 
logical in Aristotle. 


But, while philosophy is thus intimately bound up with 
positive science, it is not to be identified with it. It is 
true that in early times the distinction between the two is 
not realised. The word votya covered all we mean by 
science and a great deal more besides, such as the arts 
of making pontoons and guessing riddles. But the dis- 
tinction was there all the same. If we look at Greek 
philosophy as a whole, we shall see that it is dominated 
from beginning to end by the problem of reality (TO OP). 
In the last resort the question is always, "What is 
real ?" Thales asked it no less than Plato or Aristotle ; 
and, no matter what the answer given may be, where that 
question is asked, there we have philosophy. It is^ no 
part of the historian's task to decide whether it is a 
question that can be answered, but there is one comment 
he may fairly make. It is that the rise and progress of 
the special sciences depended, so far as we can see, on its 
being asked. We find that every serious attempt to 
grapple with the ultimate problem of reality brings with it 
a great advance in positive science, and that this has 

1 This requirement of Greek scientific method is often^ ignored, but 
Milton's Raphael knows all about it. See Paradise Lost, viii. 8 1 : "how 
build, unbuild, contrive To save appearances." 


always ceased to flourish when interest in that problem 
was weak. That happened more than once in the history 
of Greek philosophy, when the subordinate problems of 
knowledge and conduct came to occupy the first place, 
though at the same time it was just the raising of these 
problems that did most to transform the problem of 
reality itself. 

And this helps to explain why philosophy cannot be 
simply identified with science. The problem of reality, 
in fact, involves the problem of man's relation to it, which 
at once takes us beyond pure science. We have to ask 
whether the mind of man can have any contact with reality 
at all, and, if it can, what difference this will make to his 
life. To anyone who has tried to live in sympathy with 
the Greek philosophers, the suggestion that they were 
" intellectualists " must seem ludicrous. On the contrary, 
Greek philosophy is based on the faith that reality is 
divine, and that the one thing needful is for the soul, 
which is akin to the divine, to enter into communion 
with it. It was in truth an effort to satisfy what we 
call the religious instinct. Ancient religion was a some- 
what external thing, and made little appeal to this except 
in the "mysteries," and even the mysteries were apt to 
become external, and were peculiarly liable to corruption. 
We shall see again and again that philosophy sought to 
do for men what the mysteries could only do in part, 
and that it therefore includes most of what we should now 
call religion. 

Nor was this religion a quietist or purely contemplative 
one, at least in its best days. The mysteries had under- 
taken to regulate men's lives, and philosophy had to 
do the same. Almost from the beginning it was regarded 
as a life. It was no self-centred pursuit of personal 
holiness either. The man who believed he had seen the 
vision of reality felt bound to communicate it, sometimes 
to a circle of disciples, sometimes to the whole human 
race. The missionary spirit was strong from the first. 
The philosopher believed that it was only through the 


knowledge of reality that men could learn their own 
place in the world, and so fit themselves to be fellow- 
workers with God, and believing this he could not rest 
till he had spread the knowledge of it to others. The death 
of Sokrates was that of a martyr, and " intellectualism/' if 
there is such a thing, can have no martyrs. 





I. Though neither the time nor the milieu can explain 
the rise of so personal a thing as philosophy, they may 
have considerable influence on the form it assumes. It is 
not, therefore, without interest to observe that Miletos, 
"the pride of Ionia/' 1 is just the place where the con- 
tinuity of prehistoric Aegean civilisation with that of later 
times is most strongly marked. The Milesians them- 
selves believed their city to be a Cretan colony, and this 
belief has received remarkable confirmation from recent 
excavations. We now know that the old town of Miletos 
belonged to the last period of the Late Minoan civilisation, 
and that here at least that civilisation passed by imper- 
ceptible gradations into what we call the Early Ionic. 
There is a Milatos in Crete as well as in Ionia, and the 
name of Thales is at home in the island too. 2 We 
may perhaps infer that the greatness of Miletos was 
in some measure due to its inheritance from that earlier 
age which has so recently become known to us. The 
Milesians kept in close touch with Egypt and the 
peoples of Asia Minor, especially the Lydians, and their 
colonial empire extended to the northern coasts of the 

1 Herod, v. 28 : rrjs 'Iam7?s fjv TT/JOCTX^CC. 

2 See my paper, " Who was Javan ? " (Proceedings of the Classical 
Association of Scotland, 1912, pp. 91 fE). 



2. There is no reason to doubt that Thales was the 
founder of the Milesian school of cosmologists, and to all 
appearance he was the first human being who can rightly 
be called a man of science. The distinction between 
cosmologies such as the Milesian and cosmogonies such 
as that of Pherekydes is a fundamental one, and it is 
far more important to observe the points in which the 
Milesians differed from their predecessors, whether Greek 
or barbarian, than to look for survivals of primitive belief 
in their speculations. No doubt these exist, and there 
may well have been more of them than we know ; but 
for all that it is true to say that with Thales and his 
successors a new thing came into the world. 

Of Thales himself we know a great deal less than 
we should like to know. In popular tradition he lived 
mainly as one of the " Seven Wise Men," and many tales 
were told of him. In one of these he is the type 
of the unpractical dreamer, and falls into a well while 
star-gazing ; in another he shows himself superior to 
the ordinary practical man by the use he makes of his 
scientific knowledge. He is said to have foreseen an 
abundance of olives and made a corner in oil, thus prov- 
ing he could be rich if he liked. It is plain that people in 
general had no idea of his real work, and regarded 
him simply as a typical " sage," to whose name anecdotes 
( originally anonymous might be attached. These stories, 
then, tell us nothing about Thales himself, but they do 
bear witness to the impression produced by science and 
scientific men when they first appeared in a world that 
was half inclined to marvel and half inclined to scoff. 

There is, however, another set of traditions about 
Thales from which something may be learnt. They are 
not of a popular character, since they attribute to him 
certain definite scientific achievements. One of the most 
important of these, the prediction of a solar eclipse, is 
reported by Herodotos (i. 74). The existence at Miletos 
of a continuous school of cosmologists makes the pre- 
servation of such traditions quite easy to understand. 


As, however, Thales does not appear to have written 
anything, it cannot be said that our evidence is complete. 
What makes strongly in its favour is that the discoveries 
and other achievements ascribed to him are for the most 
part just such developments of Egyptian and Babylonian 
" science " as we should expect to find. But even if the 
evidence is considered insufficient, it makes little differ- 
ence. In that case Thales would become a mere name 
for us, but it would still be certain that his immediate 
successors laid the foundations of rational science. There 
can be no harm, therefore, in mentioning some of these 
traditions and interpreting them partly in the light of 
what went before and partly in that of what came after. 

3. We learn, then, from Herodotos 1 that the life of 
Thales belonged to the reigns of Alyattes and Croesus, 
kings of Lydia, and that he was still living shortly before 
the fall of Sardeis in 546 B.C. We are also told that 
at an earlier date he had predicted an eclipse of the sun 
which put an end to a battle between the Lydians and the 
Medes. That was on May 28th (O.S.), 585 B.C. Now 
there is nothing at all incredible in the story of this pre- 
diction, though it is quite certain that the true cause of 
eclipses was not discovered till after the time of Thales, 
and his successors gave quite erroneous and fantastic 
accounts of them. The Babylonians, however, were 
'equally ignorant on the subject, and yet they predicted 
eclipses with tolerable accuracy by means of a cycle of 
223 lunations. It is not even necessary to suppose that 
Thales had to visit Babylon to learn as much as this. In 
Hittite times Mesopotamian influence had been strong in 
Asia Minor, and Sardeis has been called an advanced post 
of Babylonian civilisation. There may well have been 
" wise men " in Lydia who had preserved the old secret. 
It is interesting to note also that the Lydian king seems 
to have employed the Milesian as his scientific expert ; 
for we are told that Thales accompanied Croesus on the 
expedition that proved fatal to his monarchy, and that he 
1 References to authorities are given in E. Gr. Ph? 2-7. 


diverted the course of the river Halys for him. We 
know, lastly, from Herodotos that he took a prominent 
part in politics, and that he tried to save Ionia by urging 
the twelve cities to unite in a federal state with its capital 
at Teos. 

4. We are further told on the authority of Aristotle's 
disciple Eudemos, who wrote the first history of mathe- 
y matics, that Thales introduced geometry into Hellas. It 
is extremely probable that he had learnt in Egypt the 
elementary rules of mensuration referred to in the Intro- 
duction ; but, if we may trust the tradition, he must have 
advanced beyond his teachers. He is said to have taught 
the Egyptians how to measure the height of the pyramids 
by means of their shadows, and also to have invented a 
method of finding the distance of ships at sea. It was 
common knowledge among the peoples of the East that a 
triangle whose sides were as 3 : 4 : 5 had always a right 
angle, and right angles were laid out by means of this 
triangle. What we are told of Thales suggests that he 
invented some further applications of this primitive piece 
of knowledge, and if so that was the beginning of rational 
science. At any rate, there is no reason to doubt that he 
was the pioneer of those investigations which were to bear 
fruit later in the hands of Pythagoras, though it is hardly 
safe to say more. 

5. According to Aristotle, Thales said that the earth 
floats on the water, and he doubtless thought of it as a 
t flat disc. That, at least, was the view of all his suc- 
cessors except Anaximander, and it remained characteristic 
of Ionic as distinct from Italic cosmology down to the 
time of Demokritos. It sounds primitive enough, but in 
reality it marks a notable advance. The whole history of 
cosmology at this date is the story of how the solid earth 
was gradually loosed from its moorings. Originally sky 
and earth were pictured as the lid and bottom of a sort of 
box ; but from an early date the Greeks, as was natural 
for them, began to think of the earth as an island sur- 
rounded by the river Okeanos. To regard it as resting 


on the water is a further step towards a truer view. It 
was something to get the earth afloat. 

This was no doubt connected with what Aristotle 
regards as the principal tenet of Thales, namely, that 
everything is made put of water, or, as he puts it in his 
own terminology, that water is the material cause of all 
things. We have no trustworthy information about the 
grounds on which this doctrine was based ; for, in the 
absence of any writings by Thales himself, Aristotle can 
only guess, and his guesses are apparently suggested by 
the arguments used in support of a similar theory at a 
later date. We are perhaps justified in interpreting it 
rather in the light of the doctrines afterwards held by the 
Milesian school, and especially by Anaximenes ; and, if 
we try to do this, our attention is at once called to the 
fact that in these days, and for some time after, cc air " 
(a>7/>) was identified with water in a vaporous state. In 
fact it was regarded as only a purer and more transparent 
form of mist, while a still purer form was " aether" 
(aiOyp), which is properly the bright blue of the Mediter- 
ranean sky, and is fire rather than air. It was also 
believed that this fire and that of the heavenly bodies was 
fed by vapour rising from the sea, a view which, on these 
presuppositions, is the natural one to take of evaporation. 
On the other hand, we see that water becomes solid when 
it freezes, and Anaximenes at least held that earth and 
stones were water frozen harder still. It may well have 
seemed to Thales, then, that water was the original thing 
from which fire on the one hand and earth on the other 
arose. That, of course, is a more or less conjectural 
account ; but, if Anaximenes was in any sense his 
follower, the views of Thales must have been something 
like this. His greatness, however, would lie in his having 
asked the question rather than in the particular answer he 
gave it. Henceforth the question whether everything can 
be regarded as a single reality appearing in different forms 
is the central one of Greek science, and the story we have 
to tell is how that in time gave rise to the atomic theory. 


6. The next generation of the Milesian school is 
represented by Anaximander. 1 We are on surer ground 
with regard to his doctrines ; for he wrote a book which 
was extant in the time of Theophrastos and later. It is 
probable that it was the first Greek book written in prose, 
and it may be noted here that Ionic prose was the regular 
medium of philosophical and scientific writing. Two 
Greek philosophers, Parmenides and Empedokles, wrote 
in verse at a later date, but that was quite exceptional, 
and due to causes we can still to some extent trace. 
Anaximander was also the first cartographer, and this 
connects him with his younger fellow-citizen Hekataios, 
whose work formed, as has been said, the text of Anaxi- 
mander's map. 

Anaximander seems to have thought it unnecessary to 
fix upon " air," water, or fire as the original and primary 
^form of body. He preferred to represent that simply as 
a boundless something (a-rreLpov) from which all things 
arise and to which they all return again. His reason for 
looking at it in this way is still in part ascertainable. It is 
certain that he had been struck by a fact which dominated 
all subsequent physical theory among the Greeks, namely, 
that the world presents us with a series of opposites, of 
which the most primary are hot and cold, wet and dry. 
If we look at things from this point of view, it is more 
natural to speak of the opposites as being "separated out" 
from a mass which is as yet undifferentiated than to make 
any one of the opposites the primary substance, Thales, 
Anaximander seems to have argued, made the wet too 
important at the expense of the dry. Some such thought, 
at any rate, appears to underlie the few words of the 
solitary fragment of his writing that has been preserved. 
He said that things "give satisfaction and reparation to 
one another for their injustice, as is appointed according 
to the ordering of time." This conception of justice and 
injustice recurs more than once in Ionic natural philo- 
sophy, and always in the same connexion. It refers to 

1 References to authorities are given in JR. Gr. Ph? g I ^ sqq. 


the encroachment of one opposite or " element " upon 
another. It is in consequence of this that they are both 
absorbed once more in their common ground. As that is 
spatially boundless, it is natural to assume that worlds 1 
arise in it elsewhere than with us. Each world is a sort 
of vortex in the boundless mass. Our authorities attribute 
this view to Anaximander, and no good reason has been 
given for disbelieving them. It is obviously an idea of 
the greatest scientific importance ; for it is fatal, not only 
to the theory of an absolute up and down in the universe, 
but also to the view that all heavy things tend to the same 
centre. It was, in many ways, a misfortune that Plato 
was led to substitute for this old doctrine the belief in a 
single world, and thus to prepare the way for the 
reactionary cosmology of Aristotle. The Epicureans, who 
took up the old Ionic view at a later date, were too 
unscientific to make good use of it, and actually combined 
it with the inconsistent theory of an absolute up and 
down. We are told that Anaximander called his in- 
numerable worlds u gods." The meaning of that will 
appear shortly. 

7. The formation of the world is, of course, due to the 
Cc separating out " of the opposities. Anaximander' s view 
of the earth is a curious mixture of scientific intuition and 
primitive theory. In the first place, he is perfectly clear 
that it does not rest on anything, but swings free in space, 
and the reason he gave was that there is nothing to make 
it fall in one direction rather than in another. He inferred 
this because, as has been observed, his system was incom- 
patible with the assumption of an absolute up and down. 
On the other hand, he gives the earth a shape intermediate 
between the disc of Thales and the sphere of the Pythagor- 
eans. He regarded it as a short cylinder "like the drum of 

1 1 do not use the term " world " for the earth, but as the equivalent 
of what was called an ou/>avos at this date, and later a KOCT/AOS. It means 
everything within the heavens of the fixed stars. From our point of 
view, it is a " planetary system," though the earth and not the sun is its 
centre, and the fixed stars are part of it. 


a pillar," and supposed that we are living on the upper 
surface while there is another antipodal to us. His theory 
of the heavenly bodies shows that he was still unable to 
separate meteorology and astronomy. So long as all <c the 
things aloft" (ra jmereoopa) are classed together, that is 
inevitable. Even Galileo maintained that comets were 
atmospheric phenomena, and he had far less excuse for 
doing so than Anaximander had for taking the same view 
of all the heavenly bodies. Nor was his hypothesis 
without a certain audacious grandeur. He supposed that 
the sun, moon, and stars were really rings of fire surround- 
ing the earth. We do not see them as rings, however, 
because they are encased in " air " or mist. What we do 
see is only the single aperture through which the fire 
escapes " as through the nozzle of a pair of bellows." 
We note here the beginning of the theory that the 
heavenly bodies are carried round on rings, a theory 
which held its ground till Eudoxos replaced the rings 
by spheres. We are also told that Anaximander noted 
the obliquity of these rings to what we should call the 
plane of the equator. Eclipses were caused by stoppages 
of the apertures. 

8. With regard to living beings, Anaximander held 
that all life came from the sea, and that the present forms 
of animals were the result of adaptation to a fresh environ- 
ment. It is possible that some of his biological theories 
were grotesque in detail, but it is certain that his method 
was thoroughly scientific. He was much impressed by 
the observation of certain viviparous sharks or dogfish, 
and evidently regarded them as an intermediary between 
fishes and land animals. His proof that man must have 
been descended from an animal of another species has a 
curiously modern ring. The young of the human species 
require a prolonged period of nursing, while those of 
other species soon find their food for themselves. If, 
then, man had always been as he is now he could never 
have survived. 

9. The third of the Milesians was Anaximenes, whose 


activity seems to fall in the period when Ionia had come 
under Persian rule. 1 He too wrote a prose work of which 
one fragment survives. He was not a great original 
genius like Anaximander, and in some respects his cosmo- 
logy falls far short of his predecessor's. His title to 
remembrance is really based on his discovery of the 
formula which for the first time made the Milesian theory 
coherent, that of rarefaction and condensation. He re- 
garded "air" the air we breathe, but also that which 
thickens into mist and water as the primary form of 
body, and so far his theory resembled that we have 
ascribed to Thales. On the other hand, he thought of 
this air as boundless and as containing an infinite number 
of worlds, in this respect following Anaximander. The 
solitary fragment quoted from his work shows that he was 
influenced by the analogy of the microcosm and the 
macrocosm. " As our soul," he says, " which is air, holds 
us together, so do breath and air encompass the whole 
world." The world is thought of as breathing or inhaling 
air from the boundless mass outside it, This Air he spoke 
of as a " god." 

The cosmology of Anaximenes was reactionary in many 
ways. It was felt, no doubt, that Anaximander had gone 
too far, though we shall see that his audacities contained 
the promise of the future. According to Anaximenes, the 
'earth is flat and floats upon the air "like a leaf." The 
heavenly bodies also float on the air. Their paths are not 
oblique, but the earth is tilted up, so that most of them 
are hidden when they get behind the higher side of it. It 
is unfortunate that Anaximenes did not know the spherical 
shape of the earth ; for this line of thought might have led 
him to discover the inclination of its axis. As it was, he 
regarded it as a disc, and said the heavens surrounded it 
<c like a hat." Ionia was never able to accept the scientific 
view of the earth, and even Demokritos continued to 
believe it was flat. The suggestive theory of Anaximander 
was to be developed in another region. 

1 References to authorities are given in E, Gr. PL Z 23 sqq. 


TO. It has recently been maintained that the Milesian 
cosmology was based on the primitive and popular theory 
of " the four elements." It is not meant, of course, that 
the scientific conception of an cc element " existed at this 
date. We shall see later that this was due to Empedokles, 
and it is only the place that the old quaternion of Fire, 
, Air, Earth, and Water occupied in his system, and after- 
Awards in that of Aristotle, that has led to these being 
called " the four elements." It is an unfortunate con- 
fusion, but it is very difficult to avoid it, and we must 
perforce continue to use the word "element" in two 
senses which have very little to do with one another. It 
is undeniable that, from an early date, a fourfold or three- 
fold division of this kind was recognised. It can be traced 
'in Homer and Hesiod, and it has been plausibly suggested 
that it is connected with the myth of the " portions " 
(fjLoipai) assigned to Zeus, Poseidon," and Hades. We are 
tempted, then, to say that the early cosmologists simply 
took one of these " portions " after the other and regarded 
it as primary. But, when we look closer, we shall be 
more inclined to conclude that the originality of these men 
consisted precisely in their ignoring the old popular view 
completely. In particular, we hear nothing whatever of 
earth as a primary form of body, though earth is never 
passed over in any popular list of so-called " elements." * 
This is still more striking if we remember the importance 
of Mother Earth in early cosmogonies, an importance 
which she still retains in Pherekydes. Here once more 
the breach between the Milesian cosmology and every- 
thing that had gone before is really the striking thing 
about it. 

Indeed, if we take a tjroad view of it, we shall see that 
it depends on the extension of the observed identity of 
ice, water, and steam to earth and stones on the one hand, 
and to air and fire on the other. In other words, it sub- 

1 This is pointed out by Aristotle, Met* A, 8. 989 a, 5 sqq. Neither 
he nor Theophrastos made an exception of Xenophanes. Cf, Diels, 
p. 52, 28. 


stitutes for the primitive " four elements " something which 
bears a much closer resemblance to what are now called 
the three states of aggregation, the solid, the liquid, and the 
gaseous. At any rate, the Milesians beljpved that what 
appears in these thi^ee forms was one thing, and this, as I 
hold, they called (pva-t?. 1 That term meant originally the par- 
ticular stuff of which a given thing is made. For instance, 
wooden things have one <^W, rocks another, flesh and 
blood a third. The Milesians asked for the <pvo-i? of all 
things. Thales said it was water, and we cannot be far 
wrong in guessing that he said so because, as we should 
put it, the liquid state is intermediate between the solid 
and the gaseous, and can therefore pass easily into either. 
Anaximander preferred to leave his Boundless as some- 
thing distinct from any special form of body, so that the 
opposites might proceed from it. Anaximenes saw that, 
after all, the primary substance must have some character 
of its own, and identified it with " air," that is, with the 
intermediate stage between water and fire. This he was 
able to do because he had introduced the idea of rarefac- 
tion and condensation, which alone makes the whole 
theory intelligible. In a word, the Milesians had drawn 
the outlines of the theory of matter in the physicist's sense 
of the word, and these outlines still survive in a recog- 
nisable form in our text-books. That, and not the particular 
astronomical doctrine they taught, is the central thing in 
the system, and that is why it is reckoned as the beginning 

1 Plato, Laws, 891 c : /avSweva ya/> o Xtyuv ravra irvp KCU v$up KOI 
yfjv KOI depa Trpcora rjyticrQat TCUV TTCOVTOV eTvat, /ecu r>)v <f>v(riv oi/ojLtafeiv 
ravra aura. The question really is whether the original meaning of 
<im$ is " growth." Aristotle (Met. A, 4, 1014 b, 16) did not think so ; 
for he says that, when it means " growth," it is as if one were to pro- 
nounce it with a long v. In other words, it did not at once suggest to 
him the verb <uo/xcu (Aeol. <t>viop,a.i). For controversy on this subject, 
see Heidel, ILepl <ucrea>s (Proceedings of the American Academy of Arts and 
Sciences, xlv. 4), and Lovejoy, "The Meaning of Averts in the Greek 
Physiologers " (Philosophical Review, xviii. 4). To my mind the fact that 
the Atomists called the atoms <vons is conclusive. See Ar. Phys. 265 b, 
25 ; SimpL Phys. p. 1318, 34. Atoms do not "grow." 


of philosophy. It is the earliest answer to the question, 
"What is reality?" 

The Milesian school doubtless came to an end with the 
fall of Miletos in 494 B.C., but we shall see, later that 
cc The Philosophy of Anaximenes," as it was called, con- 

tinued to be taught in other Ionian cities, and that it 
regained its influence when Ionia was once more freed 
from a foreign yoke. For the present, however, what we 
have to consider is the effect on philosophy of the Persian 

-. conquest of the Hellenic cities in Asia. 

The Breakdown of Ionian Civilisation. 

11. The spirit of Ionian civilisation had been thor- 
oughly secular, and this was, no doubt, one of the causes 
4 that favoured the rise of science. The origin of this 
secular spirit is to be found in the world described by 
Homer. The princes and chiefs for whom he sang must 
have been completely detached from the religious ideas 
which we may infer from the monuments to have been 
potent forces in the earlier Aegean civilisation. It cannot 
be said that the Olympian gods are regarded with reverence 
in the Iliad, and sometimes they are not treated seriously. 
They are frankly human, except that they are immortal 
and more powerful than men. To the religious conscious- 
ness the word "god" (0fe) always means an object of 
worship, and this is just what distinguishes the gods from 
other immortal and powerful beings (Sal/moves), In Homer, 
however, the distinction is obscured. It is by no means 
clear that all the gods in the ///Ware thought of as objects 
of worship, and it is only to a certain number of them that 
prayers and sacrifices are actually offered. It is very sig- 
nificant that when Achilles does pray in dead earnest, it is 
not to the ruler of Ida or Olympos he turns, but to the 
far-off Pelasgic Zeus of Dodona. 

The spirit of Hesiod is very different no doubt ; for he 
is no Ionian, and he feels himself to be in opposition to 
Homer, but the influence was too strong for him. He 


really did even more than Homer to dissociate the idea of 
god from that of worship. It is certain that many of the 
cc gods " in the Theogony were never worshipped by anyone, 
and some of them are mere personifications of natural 
phenomena, or even of human passions. For our present 
purpose, it is of most importance to observe that it was 
just this non-religious use of the word "god" which 
made it possible for the Milesians to apply it to their 
primary substance and their cc innumerable worlds." That 
way of speaking does not bear witness to any theological 
origin of Greek science, but rather to its complete inde- 
pendence of religious tradition. No one who has once 
realised the utterly secular character of Ionian civilisation 
will era: be tempted to look for the origins of Greek philo- 
sophy in primitive cosmogonies. "* 

12. The feudal society pictured for us >y Homer 
had been replaced in the Ionic cities by a commercial 
aristocracy, but the rhapsodes still recited Homer in the 
market-place, as the bards had done at the feudal prince's 
board. It was impossible to get away from the humanised 
Olympian gods, and in practice it was of these that men 
thought when they worshipped at the shrines founded in 
earlier days, when the gods were still awful beings to 
be approached with dread. A people brought up on 
Homer could hardly think of the gods as moral beings, 
though they were supposed to be the guardians of morality. 
Almost the only divine attribute they possessed was power, 
and even that is retained chiefly as a foil to human 
impotence, a thing of which the lonians are deeply con- 
scious. The generations of men pass away like the leaves 
of the forest, and there is no life to come, or at best a 
shadowy one, of which the departed "soul" is itself 
unconscious. Only so much is left of it as will serve to 
explain dreams and visions ; the man himself is gone 
for ever when he dies. So it is wise for men to think 
only mortal thoughts (avQp&Triva (ppovetv). The mysterious 
power that awards happiness and misery in this life, and is 
as often called <c the godhead " (TO detov) as God, appears 


to be jealous of man, and brings low everyone that exalts 
himself. So we should eat, drink, and be merry, but 
take heed withal to do ct naught too much " (wev 
ayav). The man who observes the precept "Know 
thyself" will not be puffed up. For overmuch prosperity 
(6'A/3o?) brings satiety (KO^OO?), which begets pride (Sfipify 
and that in turn the blindness of heart (arV), which God 
sends on those he is resolved to ruin. A like doctrine 
appears in the Hebrew Wisdom literature some genera- 
tions later. 

13. Such a view of life comes naturally to the 
wealthier classes in an over-civilised nation like the Ionia 
of the seventh and sixth centuries B.C., but it can bring 
no satisfaction to the people, which always demands- some 
definite satisfaction for its religious instincts. We can still 
see clear traces of a very different attitude towards the 
gods even among the lonians themselves. The Homeric 
Hymn to Apollo is, no doubt, sufficiently I secular in tone, 
but the sanctuary of Delos still retained some memories 
of the old Aegean religion. It is not for nothing that the 
boat, which in prehistoric times had conveyed the " twice 
seven " Ionian youths an^ maidens from Athens to Crete, 
went to Delos instead in later days^ and the legend 
of the Hyperboreans cdnnected Delos with still more 
remote and wonderful regions. It was not, however, 
in Ionia itself that these germs were to fructify ; for the 
days of Ionian freedom were almost at an end, and the 
citizens of one state after another had to seek new hpmes 
in the far west. A new age had begun in whidr'there 
was no room for the light-hearted polytheism of Homer. 
When men once more felt a real need of worship, that 
could not satisfy them. It is easier to worship a tree 
or an animal, than a god who is just a man freed from 
the restraints that keep ordinary men in check. That 
is also why the worship of two agricultural gods, who are 
almost unknown to Homer, Demeter and Dionysos, 
come to be of such importance at this date. They had 
not been completely humanised yet, though we can see 


the beginnings of the process in the Homeric Hymns, so 
it wad still possible for men to worship them sincerely. 


14. The cult of Dioryysos, in particular, had received 
a new impulse from the similar Thracian and Phrygian 
worships of Zagreus and Sabazios. The phenomenon of 
" ecstasy," which was prominent in all these, suggested 
an entirely different view of the soul and its relation to 
the body from that we find in Homer, and this was 
propagated by the Orphic religion, which we now find 
spreading in every direction. It was distinguished from 
all earlier Greek religion in two important respects. In 
the first place, it appealed to a revelation which had 
been written down in sacred books, and in the second 
place, it was organised in communities not based on a real 
or fictitious tie of blood, but open to all who became 
initiated and promised to obey the rule. Its teaching was 
the exact opposite of the Ionian pessimism, which had 
widened the gulf between its humanised gods and man 
so far that religion in any real sense had become impossible. 
The Orphics taught, on the contrary, that, though men 
were certainly fallen, they were yet akin to the gods 
and might rise again by a system of u purifications " 
(Ka6apiJ.ol} ; they might win " redemption " (AiW) from 
^sin and death, and dwell with the gods for evermore. For 
the soul of the Orphic " saint " (o<no?) was immortal ; 
it had existed before his birth, and would exist after 
his death. Indeed, these words are improperly used. 
What men call life is really death, and the body is 
the tomb of the soul (crS)^a cnj/xa), which is imprisoned 
successively in animal, and even in vegetable bodies, until 
its final purification liberates it from the "wheel of birth." 
Those souls, on the other hand, which are incurable 
(ai/77/cecnro*, ana-rot) are condemned to lie in the " Slough " 
{JSopfiopos) for ever. The ideas of heaven and hell, salva- 
tion and damnation, were a new thing in Greek religion. 


The Orphic religion was mainly the faith of obscure 
people. We do not know the names of its preachers and 
missionaries, and we only know it to have been a reality 
from certain gold plates buried with believers in South 
Italy and Crete. It is true that rulers like Peisistratos 
took up the religion of Orpheus for political reasons ; but, 
on the whole, it is for us anonymous. That it was apt to 
degenerate into a mere superstition is natural ; for there 
were no great Orphic teachers, so far as we know, who 
could have preserved its purity, and it fell an easy prey to 
charlatans and impostors. We shall see, however, that 
certain elements, which seemed to have permanent value, 
were taken up by the philosophers, and so preserved to 
later ages. In this way Orphicism has profoundly affected 
all subsequent religions and philosophies, and- not least 
those which seem, at first sight, to be furthest removed 
from it. 


15. It need hardly be said that such ideas were 
wholly foreign to the enlightened men of the Ionian cities. 
The saying that "all things are full of gods" is attributed 
to Thales, and belongs in any case to this period. The 
tendency it indicates is what we should call pantheistic, in 
the sense in which pantheism has been called " a polite 
atheism." This is still plainer in another form of the 
same saying, which is ascribed to Herakleitos. He asked 
his visitors to come into the kitchen, saying "Here too 
are gods." But the true spirit of Ionian science is best 
seen in some of the writings ascribed to Hippokrates, 
which are certainly not later than the fifth century B.C. 
In the treatise on The Sacred Disease (epilepsy) we 

" I do not think that any disease is more divine or more 
sacred than others. ... I think that those who first called 
this disease sacred were men such as there are still at the 
present day, magicians and purifiers (/caSa/ora/) and charlatans 
and impostors. They make use of the godhead (TO Oetov) to 
cloak and cover their own incapacity." 


And again in the treatise on Airs, Waters and Sites 

" Nothing is more divine or more human than anything 
else, but all things are alike and all divine." 

That is the true note of " enlightenment," and it was the 
note of all the Ionian schools. It is most strongly marked 
in an elegiac and satirical poet, who approached the 
question from the standpoint of the reformer rather than 
of the scientific investigator. I refer to Xenophanes, who 
is often regarded as the founder of the Eleatic school, a 
point we shall return to later. In any case, chronological 
and other considerations make it most instructive to take 
him up at this point in our story. 

1 6. It is difficult to determine the dates of Xeno- 
phanes' life with any accuracy ; for those given by ancient 
authorities have been arrived at by a mere process of com- 
bination. 1 The facts of his life are also obscure. There 
is not the slightest evidence that he was a rhapsode, and it 
is most improbable. He may have visited Elea as well as 
other places, but no ancient authority states unambiguously 
that he did. He was certainly a citizen of Kolophon, and 
we know from his own statement that he had lived in exile 
from the age of twenty-five, and that he was still writing 
poetry when he was ninety-two. There is no doubt that 
he lived chiefly in Sicily, and it is practically certain that 
he was at the court of Hiero of Syracuse, who reigned 
from 478 to 467 B.C. HeXs also said to have been a 
disciple of Anaximander, and there are features in his 
poetry which make this probable. On the whole, it is 
safe to say that Xenophanes belongs mainly to the sixth 
Century B.C., though he lived well into the fifth. Hera- 
kleitos already speaks of him in the past tense, and couples 
his name with that of Hekataios. 

17. If we look at the very considerable remains of 
his poetry that have come down to us, we shall see that 
they are all in the satirist's and social reformer's vein. 
There is one dealing with the management of a feast, 

1 References to authorities are given in . Gr. PL 2 55 Jff. 



another which denounces the exaggerated importance 
attached to athletic victories, and several which attack the 
humanised gods of Homer. 1 The problem is, therefore, 
to find, if we can, a single point of view from which all 
these fragments can be interpreted. It may be that no 
such point of view exists ; but, if one can be found, it is 
likely that we shall understand Xenophanes better. Now 
we know that a great change came over Hellenic life at 

* the end of the sixth century B.C. It was a reaction against 
the somewhat effeminate refinement and daintiness 
(a/fycmy?) of Ionia, which had its source in the court of 
Sardeis and had spread with Ionian colonisation even to 
the far West. It had reached its highest point at the 
court of Polykrates of Samc^s, and its singers were 
Mimnermos of Kolophon and x Anakreon of Teos. It was 

j.not coarse and brutal like the luxury of later days, but 
there was an element of decadence in it. It was charac- 
terised fi once by pessimism and frivolity. The change 
came when " the Mede appeared" (Xenophanes, fr. 22), 
and the lonians had no longer to do with half-Hellenised 
Lydians, but with a sterner foe. They then began to feel 
the gulf that divided the Hellene from the " barbarian," 
and to accentuate the differences between them more and 
more. The general use of the name " Hellenes " dates 

tpnly from this time. Thucydides (i. 6) notes the change 
in dress which marked the new spirit, and his statement 
is confirmed by vase-paintings. 2 In architecture the Doric 

.style supersedes the Ionic. Everywhere we note a return 
to a simpler and more virile way of life. It seems to me 
that Xenophanes is best understood as a pioneer of this 
movement. 8 

1 8. The religious reformers of the day turned their 
back on the anthropomorphic polytheism of Homer and 
Hesiodj and Xenophanes will have none of it either. Tn 

1 For a translation of the fragments, see JE. Gr. ?h* 57. 

2 Sec Pernice in Gercke and Norden's Einkitung, vol. ii. pp. 39-44. 
a See especially fr. 3, 


his case, however, this revolt is based on a conviction that 
the tales of the poets are directly responsible for the 
- moral corruption of the time. " Homer and Hesiod 
have ascribed to the gods all things that are a shame and 
a disgrace among mortals, stealings and adulteries and 
deceiving of one another" (fr. n). And this he held 
was due to the representation of the gods in human 
form. Men make gods in their own image ; those of 
the Ethiopians are black and snub-nosed, those of the 
Thracians have blue eyes and red hair (fr. 16). If horses 
or oxen or lions had hands and could produce works of 
art, they too would represent the gods after their own 
fashion (fr. 15). All that must be swept away along 
with the tales of Titans and Giants, those " figments of 
an earlier day " (fr. i) if social life is to be reformed. 

Xenophanes found the weapons he required for his 
attack on polytheism in the science of the time. There 
are traces of Anaximander's cosmology in .the fragments, 
and Xenophanes may easily have been his disciple before 
he left Ionia. He seems to x have taken the gods of 
mythology one by one and reduced them to meteoro- 
logical phenomena, and especially to clouds. And he 
maintained there was only one god namely, the world. 
That is not monotheism, as it has been called, but pan- 
theism. It is a simple -reproduction of that special use 
of the term " god " we have seen to be characteristic 
of the early cosmologists generally. There is no evidence 
that Xenophanes regarded this cc god " with any religious 
feeling, and all we are told about him (or rather about it) 
is purely negative. He is quite unlike a man, and has no 
special organs of sense, but " sees all over, thinks all 
L over, hears all over " (fr. 24). Further, he does not go 
about from place to place (fr. 26), but does everything 
" without toil " (fr. 25). It is not safe to go beyond this ; 
for Xenophanes himself tells us no more. It is pretty 
certain that if he had said anything more positive or more 
^definitely religious in its bearing it would have been 
quoted by later writers. 


19. But while Xenophanes makes use of contem- 
porary science to overthrow the Olympian hierarchy, it is 
plain that he was not himself a scientific man. In spite 
of Anaximander, he still believes in a flat earth extending 
to infinity in all directions, and boundless in depth also. 
Consequently it is a different sun that traverses our 
heaven every day. The same must apply to the moon, 
which he further held to be superfluous. Both sun and 
moon are ignited clouds. The stars, too, are clouds that 
go out in the day time, but glow at night like charcoal 
embers. That is not science as science was understood 
at Miletos, and it seems that Xenophanes merely made 
use of cosmological ideas for his own purposes. Any 
stick was good enough to beat the gods of Homer and 
Hesiod with. He says distinctly that the accounts he 
gives of the gods are " guesses like the truth " (fr. 34), 
and he denies the possibility of certain knowledge in 
this field " Even if a man should chance to say the 
complete truth, he cannot know that it is the truth" 
(fr. 34). In all this Xenophanes is the precursor of 
another philosophy that came from Ionia at a later date, 
that of Epicurus. The difference is mainly that it was 
less of an anachronism in the fifth century B.C. than it was 
two hundred years later. 

In this chapter we have seen how the traditional view 
of the world broke down, and how its place was taken by 
Orphic mysticism on the one hand and by enlightened 
scepticism on the other. Neither of these contained in 
itself the promise of the future. That lay in the work of 
the man who first united science with religion, Pythagoras 
of Samos. 



The Problem 

20. Pythagoras must have been one of the world's 
greatest men, but he wrote nothing, and it is hard to say 
how much of the doctrine we know as Pythagorean is due 
to the founder of the society and how much is later 
development. 1 We have met the same difficulty in the 
case of Thales, and we shall meet it again when we come 
to Sokrates. One general remark may be made about it 
at once. So far as we know, all great advances in human 
knowledge have been due to individuals rather than to 
the collective work of a school, and so it is better to take 
the risk of ascribing a little too much to the founder than 
to lose sight of him among a crowd of disciples. On the 
other hand, it is certain that some Pythagorean doctrines 
at least belong to a later generation, and it will be well to 
reserve these for a future chapter. Such a division is 
inevitable if we are to give an intelligible account of 
Pythagoreanism, but it 'must be remembered that it is 
often quite uncertain whether a particular doctrine belongs 
to the earlier period or to the later. 

21. It is also hard to say how much of what we are 
told about the life of Pythagoras is trustworthy ; for a 

1 Aristotle never attributes any doctrine to Pythagoras himself. He 
generally speaks of " the so-called Pythagoreans," and, often, still more 
cautiously, of " some of the Pythagoreans." References to authorities 
are given in E. Gr. Ph? 37 s$$. 


mass of legend gathered round his name at an early date. 
Sometimes he is represented as a man of science, and 
sometimes as a preacher of mystic doctrines, and we 
might be tempted to regard one or other of those charac- 
ters as alone historical. It is quite possible to picture 
Pythagoras as a mere medicine-man, and to treat all 
Pythagorean science as the work of his successors. It is 
also possible to rationalise the story of his life and repre- 
sent him mainly as a mathematician and statesman. In 
that case we have to regard the miraculous tales told of 
him as due to the Neopythagoreans of the early centuries 
of our era. There is a serious difficulty here, however ; 
for many of these wonders were already known to 
Aristotle. It is equally difficult to reject the tradition 
that makes Pythagoras the true founder of mathematical 
science ; for that science was certainly in existence by the 
middle of the fifth century B.C., and it must have been the 
work of someone. If the credit is really due to another 
than Pythagoras, it is strange that his name should have 
been forgotten. Further, Herakleitos in the next genera- 
tion tells us that Pythagoras practised inquiry (broplti) 
beyond all other men, and he thinks the worse of him for 
it. That is practically contemporary evidence, and it can 
only mean that Pythagoras was famous as a man of 
science. The truth is that there is no need to reject 
either of the traditional views. The union of mathe- 
matical genius and mysticism is common enough. It was 
also characteristic of the seventeenth century, which took 
up once more the thread of Greek science, Kepler was 
led to discover the laws of planetary motion by his belief 
in the " harmony of the spheres " and in planetary souls. 

Life and Doctrine. 

22. Pythagoras was a Samian, and, as we are told, he 
migrated to Italy because he disliked the rule of Poly- 
krates. That is why his floruit is given as 532 B.C., the 
year Polykrates became tyrant. No actual dates are 


known, but it is safe to say that his activity belongs 
mainly to the last quarter of the sixth century B.C. When 
he left Samos, he founded at Kroton in southern Italy a 
society which was at once a religious community and a 
scientific school. Such a body was bound to excite 
jealousy and mistrust, and we hear of many struggles. 
Pythagoras himself had to flee from Kroton to Meta- 
pontion, where he died. The chief opponent of Pytha- 
goreanism, Kylon, is expressly said to have been rich and 
noble, and there is no evidence for the belief that Pytha- 
goras and his followers took the aristocratic side. That 
notion was based on the fancy that they represented " the 
Dorian ideal." It is far from clear what is meant by the 
Dorian ideal ; but in any case Pythagoras himself was an 
Ionian, and his society was established in Achaian, not 
Dorian, colonies. It is also certain that the earlier Pytha- 
goreans used the Ionic dialect 1 After the death of the 
Master, the disturbances went on more than ever, and 
soon after the middle of the fifth century there was a 
regular rising, in the course of which the Pythagorean 
lodges (avveSpia) were burnt down, and many of the 
brethren lost their lives. Those who survived took 
refuge at Thebes and elsewhere, and we shall hear more 
of them later. 

Being a Samian, Pythagoras would naturally be 
influenced by the cosmology of the neighbouring Miletos. 
It is stated that he was a disciple of Anaximander, which 
is no doubt a guess, but probably right. At any rate his 
astronomy was the natural development of Anaximander's 
theory of planetary rings, though it went far beyond that. 
The importance of the infinite (TO otTretpoi/) in the Pytha- 
gorean cosmology suggests Milesian influence, and the 
identification of the infinite with "air" by at least some 
Pythagoreans points to a connexion with the doctrines 

1 It has been said that the name Pythagoras is Dorian in form. 
Herodotos and Herakleitos and Demokritos call him "Pythagores," and 
so no doubt he called himself. The form "Pythagoras" is no more 
Doric than " Anaxagoras." It is simply Attic. 


of Anaximenes. The way in which the Pythagorean 
geometry developed also bears witness to its descent from 
that of Miletos. The great problem at this date was the 
duplication of the square, a problem which gave rise to the 
theorem of the square on the hypotenuse, commonly 
known still as the Pythagorean proposition (Euclid, I. 47). 
If we were right in assuming that Thales worked with the 
old 3:4:5 triangle, the connexion is obvious, and the 
very name " hypotenuse " bears witness to it ; for that 
word means the rope or cord " stretching over against'* 
the right angle, or, as we say, <c subtending " it. 

23. But this was not the only influence that affected 
Pythagoras in his earlier days. He is said to have been a 
disciple of Pherekydes as well as of Anaximander, and the 
mystical element in his teaching is thus accounted for. 
In any case, as has been indicated already, the religion of 
the Delian and Hyperborean Apollo had a mystical side. 
The legends of Abaris and Aristeas of Prokonnesos are 
enough to show that. There are several points of contact 
between this form of mysticism (which seems to be inde- 
pendent of the Dionysiac) and Crete. We have seen that 
the boat containing the seven youths and seven maidens 
went to Delos in historical times, though tradition remem- 
bered its original destination was Crete, and Epimenides, 
the great purifier, was a Cretan. There are many things, 
in fact, which suggest that this form of mysticism had 
survived from "Minoan" times, and it is therefore quite 
unnecessary to seek its origin in Egypt or India. It is 
highly probable, then, that Pythagoras brought his ascetic 
practices and mystical beliefs about the soul from his 
Ionian home, and there was a statue of Aristeas of Prokon- 
nesos at Metapontion, where Pythagoras died. This does 
not, of course, exclude the possibility that the religion of 
the Pythagoreans was also influenced by contemporary 
Orphicism ; it is only meant that they derived it from a 
genuinely Ionic source, and that Apollo, not Dionysos, 
was their special god. 

24. Now one of the leading ideas of the Apollonian 


religion which had its centre at Delos in historical times 
was purification (icddapo-i^, 1 and that held an important 
place in the teaching of Pythagoras. The longing for 
purity is something very deeply rooted in human nature, 
and Catharism is always reappearing in new forms. Of 
course we may mean very different things by purity. It 
may be merely external, and in that case it can easily be 
secured by the strict observance of certain abstinences and 
taboos. That these were observed in the Pythagorean 
society is certain, and it is quite likely that many members 
of it got no further. It is certain, however, that the lead- 
ing men of the order did. There was an important medical 
school at Kroton even before Pythagoras went there, and 
it appears that the old religious idea of purification was 
early regarded in the light of the medical practice of 
purgation. At any rate, Aristoxenos, who was personally 
acquainted with the Pythagoreans of his time, tells us that 
they used medicine to purge the body and music to purge 
the soul. That already connects the scientific studies of 
the school with its religious doctrine, since there is no 
doubt that we owe the beginnings of scientific therapeutics 
and harmonics to the Pythagoreans. But that is not all. 
In the Phaedo Sokrates quotes a saying that " philosophy 
is the highest music," which seems to be Pythagorean in 
origin. The purgative function of music was fully recog- 
nised in the psychotherapy of these days. It originated 
in the practice of the Korybantic priests, who treated 
nervous and hysterical patients by wild pipe music, thus 
exciting them to the pitch of exhaustion, which was 
followed in turn by a healthy sleep from which the patient 
awoke cured. An interesting light is thrown on this by 
what was known as " Tarantism " in later days.* Taking 
all these things together, there is much to be said for the 
view that the originality of Pythagoras consisted in this, 
that he regarded scientific, and especially mathematical, 

1 Farnell, Cults of the Greek States, vol. iv. pp. 295^. 
8 See Enc. Brit, (nth edition) J.P. "Tarantula." 


study as the best purge for the soul. That is the theory 
of the early part of Plato's Phaedo^ which is mainly a state- 
ment of Pythagorean doctrine, and it frequently recurs in 
the history of Greek philosophy. It may be added that 
tradition represents the word " philosophy " as having 
been first used by Pythagoras. If that is so (and there is 
much to be said for the tradition), we need not hesitate to 
ascribe to him the saying mentioned in the Phaedo that 
philosophy is the "highest music," and so, since music was 
certainly regarded as a soul-purge, we come to the same 
result in another way. We still speak of " pure mathe- 
matics," 1 and that way of speaking has given rise in turn 
to the phrase " pure scholarship." 

25. Closely connected with this is the doctrine of the 
Three Lives, the Theoretic, the Practical, and the Apo- 
laustic, which is probably to be referred to the founder of 
the society. There are three kinds of men, just as there 
are three classes of strangers who come to the Olympic 
Games. The lowest consists of those who come to buy 
and sell, and next above them are those who come to 
compete. Best of all are those who simply come to look 
on (OGwpeiv). Men may be classified accordingly as lovers 
of wisdom (<pi\6cro(f>oC) y lovers of honour (<jtuXo'rf,u<H), and 
lovers of gain (<f>i\oKepSety. That seems to imply the 
doctrine of the tripartite soul, which is also attributed to 
the early Pythagoreans on good authority, 2 though it is 
common now to ascribe it to Plato. There are, however, 
clear references to it' before his time, and it agrees much 
better with the general outlook of the Pythagoreans. The 
comparison of human life to a gathering (iravriyvpii) like 
the Games was often repeated in later days, 8 and is the 
ultimate source of Banyan's " Vanity Fair/' The view 

1 Cp. the use of KaOapQs yvQvai, ttSevai,, etc., in the Phaedo^ 65 e, 
66 d, e. 

2 The authority is Poseidonios. See my edition of the Phaedo, 68 c, 
2, note. 

8 Cp. Menander, fr. 481 Kock (Pickard-Cambridge, p. 141. No. 68), 
Epictetus, ii. 14, 23. 


that the soul is a stranger and a sojourner in this life was 
also destined to influence European thought profoundly. 

26. There can be no doubt that Pythagoras taught 
the doctrine of Rebirth or transmigration, 1 which he may 
have learned from the contemporary Orphics. Xenophanes 
made fun of him for pretending to recognise the voice 
of a departed friend in the howls of a beaten dog (fr. 7). 
Empedokles seems to be referring to him when he speaks 
(fr. 129) of a man who could remember what happened 
ten or twenty generations before. It was on this that the 
doctrine of Reminiscence, which plays so great a part in 
Plato's Meno and Phaedo, was based. 2 The things we 
perceive with the senses, we are told, remind us of things 
we knew when the soul was out of the body and could 
perceive reality directly. We have never seen equal 
sticks or stones, but we know what equality is, and it Is 
just by comparing the things of sense with the realities of 
which they remind us that we judge them to be imperfect. 
I see no difficulty in referring this doctrine in its mathe- 
matical application to Pythagoras himself. It must have 
struck him that the realities he was dealing with were not 
perceived by the senses, and the doctrine of Reminiscence 
follows easily from that of Rebirth. 

27. As has been indicated, there is more difficulty 
about the cosmology of Pythagoras. Hardly any school 
ever professed such reverence for its founder's authority 
as the Pythagorean. " The Master said so " (auro? e<a, 
ipse dmf] was their watchword. On the other hand, few 
schools have shown so much capacity for progress and for 
adapting themselves to new conditions. The contradic- 
tion here is doubtless more apparent than real, but it 
creates a difficulty for the historian, and we can hardly 
ever feel sure to what stage of development any given 

x The word metempsychosis is not used by good writers, and is 
inaccurate ; for it would mean that different souls entered into the same 
body. The older word is TraAtyyevecr/a, being "born again." See 
E. Gr. PL 2 p. 10 1, n. 2. 

* See my edition of the Phaedo, 72 e, 4 note. 


statement about Pythagoreanism refers. One thing, 
however, we can see distinctly. There is a form of the 
doctrine that precedes the rise of the Eleatic philosophy, 
and there is a form that is subsequent to it. We shall do 
well, therefore, to reserve for the present all doctrines 
which seem to imply the Eleatic criticism. That is really 
the only criterion we can apply. 

28. We can make out pretty clearly to begin with 
that Pythagoras started from the cosmical system of 
Anaximenes. Aristotle tells us that the Pythagoreans 
represented the world as inhaling u air " from the bound- 
less mass outside it, and this " air " is identified with cc the 
unlimited." On the other hand, Pythagoras seems to have 
learnt from Anaximander that the earth is not a flat disc. 
He still, in all probability, thought of it as the centre of 
the world, though his followers held otherwise at a later 
date, but he could no longer regard it as cylindrical. As 
soon as the cause of eclipses came to be understood, it 
was natural to infer that the earth was a sphere, and 
we may probably attribute that discovery to Pythagoras 
himself. With this exception, his general view of the 
world seems to have been distinctly Milesian in character. 

When, however, we come to the process by which 
things are developed out of the cc unlimited," we observe 
a great change. We hear nothing more of "separating 
out " or even of rarefaction and condensation. Instead of 
that we have the theory that what gives form to the 
Unlimited (airetpov) is the Limit (Tre/w). That is the 
great contribution of Pythagoras to philosophy, and we 
must try to understand it. We have seen that the 
Milesians had reached the conception of what we call 
"matter"; it was the work of the Pythagoreans to 
supplement this by the correlative conception of " form." 
As this is one of the central problems of Greek philosophy, 
it is very important for us to ascertain if we can what was 
originally meant by the doctrine of the Limit 

Now the function of the Limit is usually illustrated from 
the arts of music and medicine, and we have seen how 


important these two arts were for the Pythagoreans, so it 
is natural to infer that the key to its meaning is to be 
found in them. Let us see, then, what can be safely 
affirmed with regard to early Pythagorean musical and 
medical theory. The doctrines described in the following 
paragraphs are all genuinely Pythagorean, but it will be 
remembered that our ascription of any particular state- 
ment to Pythagoras himself is conjectural. We cannot 
tell either whether music or medicine came first, or, in 
other words, whether the purge of the body was explained 
by the purge of the soul, or vice versa. It will, however, 
be convenient to begin with music. 


29. In the first place, it may be taken as certain 
that Pythagoras himself discovered the numerical ratios 
which determine the concordant intervals of the scale. 
Of course, when the Greeks called certain intervals con- 
cordant (ovfjt.cfxjova) they were thinking primarily of notes 
sounded in succession and not simultaneously. In other 
words, the term refers to melodic progressions, and not to 
what we call harmonious chords. The principle is ulti- 
mately the same, indeed, but it is often of importance 
to remember that there was no such thing as harmony 
in classical Greek music, and that the word "harmony" 

means in the Greek language, first "tuning,'* 
and then " scale/' 

In the time of Pythagoras the lyre had seven strings, 
and it is not improbable that the eighth was added later as 
the result of his discoveries. All the strings were of 
equal length, and were tuned to the required pitch by 
tension and relaxation (eV/ra<ro, a^eo-i?). This was done 
entirely by ear, and the first thing was to make the 
two outside strings (hypate and nefe)' 1 concordant, in the 

1 Observe that the terms vTrar?? and vyrrj do not refer to pitch. As a 
matter of fact, the VTrdrrj gave the lowest note and the vyrr) the highest. 
The terms for " high " and " low " are ous (acutus, " sharp "), and J 


sense explained, with one another, with the middle string 
(mese\ and with the string just above it (trife, later 
paramese). The notes (tyQoyyoi) of these four strings 
were called "stationary" (eo-rcore?), and were similarly 
related to one another in every kind of scale ; the notes of 
the other three (or four in the eight-stringed lyre) were 
"movable" (/ai/ov/^o*), and scales were distinguished as 
enharmonic, chromatic, and diatonic (with their varieties), 
according as these strings were tuned more or less closely 
to the same pitch as the nearest fixed notes. They might 
differ from these in pitch by as little as what we call 
a quarter-tone, or as much as what we call a double tone. 
It is obvious that none of our scales could be played on a 
seven-stringed lyre at all ; an eight-stringed lyre, tuned 
to the diatonic scale, is required for them. Even in that 
scale, however, the Greeks did not recognise the interval 
we call the third as concordant. 1 

^ 30. It is quite probable that Pythagoras knew the 
pitch of notes to depend on the rate of vibrations which 
communicate " beats " or pulsations (TrA^a/) to the air. 
At any rate, that was quite familiar to his successors ; but 
neither he nor they had any means of measuring the rate 
of vibrations. As, however, the rate of vibration of two 
similar strings is inversely proportional to their length, it 
was possible for him to transform the problem and attack 
it on that side. The lyre did not immediately suggest 
this ; for its strings were of equal length, but a few 
experiments with strings of unequal length would establish 
the truth. Pythagoras doubtless used a simple appa- 
ratus, consisting of a string which could be stopped at 
different intervals by a movable bridge (the mono ^chord], and 
in this way reduced the experiment to a simple comparison 
of lengths on a single string. The result was to show 
that the concordant intervals of the scale could be expressed 

1 An elementary knowledge of the Greek lyre is essential for the 
understanding of Greek philosophy. A useful introduction to the 
subject will be found in the articles (by D. B. Monro) Lyra and Musica 
in Smith's Dictionary o 


by the simple numerical ratios 2:1, 3:2, and 4:3, 
or, taking the lowest whole numbers which have these 
ratios to one another, that the four stationary notes of the 
lyre could be expressed thus : 

6 8 9 12 

For convenience let us represent these four notes by those 
of the gamut in descending order : 

Nete Paramese Mes3 Hypate 

Mi Si La Mi, 

and we may explain the discovery of Pythagoras as follows : 

(1) When he took a length of string double that which 
gave the high Mi, it gave the low Mi. That is the interval 
which we call the octave and the Greeks called diapason 
(Sia Traa-wv, sc. x<>/o<5&>y). It is expressed by the ratio 2 : I 
(SL7r\dcrio$ Aoyos 1 ). 

(2) When he took a length of string half as long again as 
that which gave the high Mi, it gave La. That is the 
interval which we call the fifth and the Greeks called 
dia pente (Sta TreVre, sc. xoptScov). It is expressed by the ratio 
3 : 2 (rjjULtoXio? Xoyo?). 

(3) When he took a length of string one-third again as 
long as that which gave the high Mi, it gave Si. That 
is the interval which we call the fourth and the Greeks 
called diatessaron (Sia recro-apcoj/, sc. xo/xScSy). It is expressed 
by the ratio 4 : 3 (ex/rpiro? Xoyo9). 

(4) The compass (/*eye0os) of the octave is a fifth and 
a fourth (f x = V)> anc * the note which is a fifth from the 
riete is a fourth from the hypate, and vice versa. 

(5) The interval between the fourth and the fifth is 
expressed by the ratio 9 : 8 (eTroySoo? Xoyo?). This is called 
the "tone" (TOI/O?) or pitch par excellence (probably from 
its importance in attuning the two tetrachords to one another). 

(6) As there is no (numerical) mean proportional between 
I and 2, neither the octave nor the tone can be divided into 
equal parts. 

There is good reason for holding that Pythagoras did 
not go any further than this, and that no attempt was 
made to determine the ratios between the "movable" 
notes of the tetrachord till the days of Archytas and Plato. 


It is by no means clear, in fact, that there was any strict 
rule with regard to these at this date. 1 Aristoxenos tells 
us that the diagrams of the older musical theorists all referred 
to the enharmonic scale, which proceeded by what he called 
quarter-tones and a double tone ; but Pythagoras could not 
admit the possibility of quarter-tones, since the tone did 
not admit of equal division. The internal notes of the 
tetrachord must, then, have been regarded as of the nature 
of the " unlimited," and the " limit " was represented only 
by the perfect concords. 

31. Now if we look at the four terms (Spot) which 
we have discovered, we shall find that 8 and 9 are 
related to the extremes 6 and 12 as means. The term 9, 
which represents the note of the mese, exceeds and is 
exceeded by the same number, namely 3. It is what is 
called the arithmetical mean (a^0w:^ /xeo-or^?). On the 
other hand, the term 8, which represents the note of the 
paramese, exceeds and is exceeded by the same fraction of 
the extremes ; for 8 = 1 2 ^ = 6 + f. This was called 
the subcontrary (wei/aima), or later, for obvious reasons, 
the harmonic mean (apjmovuc}] jmeo-oryi). The geometrical 
mean is not to be found within the compass of a single 

Now this discovery of the Mean at once suggests a new 
solution of the old Milesian problem of opposites. We 
know that Anaximander regarded the encroachment of one 
opposite on the other as an " injustice," and he must 
therefore have held there was a point which was fair to 
both. That, however, he had no means of determining. 
The discovery of the Mean suggests that it is to be found 
in a " blend " (/cpaani) of the opposites, which might be 
numerically determined, just as that of the high and low 
notes of the octave had been. The convivial customs of 
the Greeks made such an idea natural to them. The 
master of the feast used to prescribe the proportions of 
wine and water to be poured into the mixing-bowl before 

1 See Tannery, " A propos des fragments philola'iques sur la musique " 
(Rev. de philologie, 1904, pp. 233 /<#,). 


it was served out to the guests. That is why the Demi- 
ourgos in Plato's Timaeus uses a mixing-bowl (jcpcmqp). It 
may well have seemed that, if Pythagoras could discover 
the rule for blending such apparently elusive things as 
high and low notes, the secret of the world had been 

32. There remains one point of which the full signi- 
ficance will not appear till later, but which must be men- 
tioned here. It is plain that the octachord scale could be 
increased by the addition of one or more tetrachords at 
either end, and that it would therefore be possible to 
obtain octave scales in which the smaller and larger inter- 
vals 1 occurred in a different order. We can get some 
rough idea of this by playing scales on the white notes of 
the piano alone. It is fortunately unnecessary for our 
present purpose to discuss the relation of these " figures of 
the octave" (e?oV rov Sia TracnSi/), as they were called, to 
the " modes " (apjmovlcu, rpoTroi) of which we hear so much 
in Greek writers ; for it cannot be said that this problem 
has been satisfactorily solved yet. 2 All that is important 
for us is that these scales were called " figures" (e?/) just 
because they varied in the arrangement of their parts. 
We have the authority of Aristoxenos for that, 3 and we 
shall see that it is a matter of fundamental importance. 


33. In Medicine we have also to do with " opposites," 
such as the hot and the cold, the wet and the dry, and it 

1 The example given by Aristoxenos is taken from the enharmonic 
tetrachord, in which, according to his terminology, we may have (i) 
\ tone, \ tone, ditone, (2) J tone, ditone, J tone, or (3) ditone, J tone, 
\ tone. 

2 See Monro, Modes of Ancient Greek Music (i 894) ; Macran, The Har- 
monics of Arlstoxenus (1902) \ J. D. Dennistoun, " Some Recent Theories 
of the Greek Modes" (Classical Quarterly, vii. (1913), PP- 83^). 

8 Aristoxenos, El. Harm, iii, 74, is quite clear that etfy here means 

figures," 8ia<ep 8' ^lv ovBlv $os Aeyeti/ $ rwpa.' ^popev yap 
TO, ovo/mra ITTI T& CITJTO. 


is the business of the physician to produce a proper "blend** 
(/cpao-*?) of these in the human body. In a well-known 
passage of Plato's Phaedo (86 b) we are told by Simmias 
that the Pythagoreans held the body to be strung like an 
instrument to a certain pitch, hot and cold, wet and dry 
taking the place of high and low in music. According to 
this view, health is just being in tune, and disease arises 
from undue tension or relaxation of the strings. We 
still speak of "tonics" in medicine as well as in music. 
Now the medical school of Kroton, which is represented 
for us by Alkmaion, based its theory on a very similar 
doctrine. According to him, health depended on the 
"isonomy" (tcroi/o/uiy) of the opposites in the body, and 
disease was just the undue predominance of one or the 
other. We need not be surprised, then, to find that 
Alkmaion was intimately associated with the Pythagoreans, 
and that he dedicated his medical treatise to some of the 
leading members of the society. Health, in fact, was an 
*' attunement " (djo/xowa) depending on a due blend of 
opposites, and the same account was given of many other 
things with which the physician is concerned, notably 
of diet and climate. The word " blend " (Kp&crii) itself 
was used both of bodily temperament, as we still call it, 
and of the temperature which distinguished one climate 
from another. When we speak of cc temperance " in 
eating and drinking, we are equally on Pythagorean 

Now we find the word we have translated "figure" 
(elSoi) used more than once in the literature of the fifth 
century B.C. in connexion with disease and death, and, as 
has been pointed out, 1 it occurs in many places in close 
connexion with a verb (jrafcWacrftu) which has also a 
technical sense in ancient medicine. The same verb (and 

1 See A. E. Taylor, Varia Socratica (St. Andrews University Publica- 
tions, No. ix.), p. T 89. Professor Taylor has not cited the t!2ty ro-G Sicl 
Trctcrcov in confirmation of his view, but it seems to me important, seeing 
that we have the express authority of Aristoxenos foj 1 rfoo? = cr^/xa in 


its substantive Karao-ravii) is also applied to the individual 
constitution of a given body. It is surely natural to inter- 
pret these uses of the word in the light of the " figures 
of the octave " explained above. The opposites on which 
health and disease depend may combine in various patterns, 
as it were, and such variation of pattern is also the explana- 
tion of the differences between the constitutions (/caret- 
o-rao-ei?) of individual patients. 


34. Having discovered that tuning and health were 
alike means arising from the application of Limit to the 
Unlimited, and that this resulted in the formation of 
certain " figures " (ezSg), it was natural for Pythagoras to 
look for something of the same kind in the world at 
large. The Milesians had taught that all things issued 
from the Boundless or Unlimited, though they had given 
different accounts of this. Anaximenes had identified it 
with "air," and had explained the forms this took by 
rarefaction and condensation. He was thinking chiefly 
of " air " as a form of mist. Pythagoras would seem to 
have regarded it mainly from another point of view ; for 
the Pythagoreans, or some of them, certainly identified 
" air " with, the void. This is the beginning, but no more 
than the beginning, of the conception of abstract space 
or extension, and what chiefly interested Pythagoras, so 
far as we can see, was the problem of how it became 
limited so as to present the appearance of the world we 

There is a striking confirmation of this in the Second 
Part of the poem of Parmenides, if, as we shall see 
reason for believing, that is a sketch of Pythagorean 
cosmology. There the two "forms" (popped), which 
men have erroneously assumed are Light and Darkness. 
Darkness was still regarded in these days as a thing, not 
as a mere privation of light, and "air" was very closely 
associated with it. In Plato's Timaeus (58 d) we have 


what is no doubt the traditional Pythagorean view, that 
mist and darkness were alike forms of " air." Now Light 
and Darkness are included in the famous Pythagorean 
table of " opposites," where they come under the head of 
Limit and the Unlimited respectively. 

35. Briefly stated, the doctrine of Pythagoras was 
that all things are numbers, and it is impossible for us 
to attach any meaning to this statement unless we have 
a clear idea of what he is likely to have meant by a 
" number." Now we know for certain that, in certain 
fundamental cases, the early Pythagoreans represented 
numbers and explained their properties by means of dots 
arranged in certain " figures " (eT^, o^/wmi) or patterns. 
That is, no doubt, very primitive ; for the practice is 
universal on dice and such things from the earliest times. 
The most celebrated of these Pythagorean figures was the 
tetraktysj- by which the members of the Order used to 
swear. This showed at a glance what the Pythagoreans 
conceived to be the most important property of the 
number ten namely, that it is the sum of the first four 
natural integers (1 + 2 + 3 + 4=10), thus 

It is obvious that this figure could be extended indefinitely, 
and that it takes the place of a formula for the sums of 
the series of successive natural integers, 3, 6, 10, 15, 21, 
and so on. These, therefore, were called " triangular 

We hear in the next place of square (rerpaj^voi) and 
oblong (Tpoimr}K:Li) numbers, A square number meant 
(as it still does) a number which is the product of equal 

1 For the form of this word cp. rpi/crvs (Att T/HTTVS'). The forma 
TpiKTvap)(o$ and rpiKrvap^lv occur in Delian inscriptions (Dhtenberger, 
SyU&<*, 588, 19 '??) 



factors, an oblong number, one which is the product of 
unequal factors. These may be presented thus 

We see at once from these figures that the addition of 
successive odd numbers in the form of a gnomon produces 
square numbers (4, 9, 16, etc.), while the addition 
of successive even numbers produces oblong numbers 
(6, 12, 20, etc.). We might go on in the same way to 
study the properties of cubic numbers, but we cannot tell 
how far Pythagoras had advanced in this direction. The 
important thing to notice is that all these figures express 
the sums of series of different kinds. The series of 
integers yields triangular numbers, that of odd numbers 
yields square numbers, and that of even numbers yields 
oblong numbers, Aristotle notes further that the form 
(etSoi) of the square numbers is always the same ; it is 
the ratio I : I. On the other hand, each successive oblong 
number has a different form (eiSo?). These correspond 
exactly to the concordant intervals of the octave. 1 

Our knowledge of these things comes chiefly from 
Neopythagorean writers, who regarded the cc figures " as 
more " natural " than the ordinary notation by letters of 
the alphabet, but they certainly were known to Aristotle, 2 

1 Thus the ratio between the sides of 2 (2 : l) is the StTrXacrtos Aoyos 
(the octave) ; the ratio between the sides of 6 (3 : 2) is the f)fj,io\io<s Aoyos 
(the fifth) ; the ratio between the sides of 12 (4:3) is the ITTIT/HTOS 
Aoyos (the fourth). 

2 Cp. especially Met. N, 5. 1092 b, 8 (Eurytos and ot TOUS <x/H0/xos 
ayovres ds ra o-)^/mTa rpiycovov KOI rerpaycovov). In Phys. F, 4, 
203 a, 13, in explaining square and oblong numbers, he uses the old 
word 6$os instead of the more modern cr^/ia. That et$os originally 
meant " figure" in the sense of" pattern*' appears from the use of 
for the figures on a piece of embroidery (Plut. Them. 29). 


and we need have no hesitation in referring them to the 
very beginnings of Pythagorean science. In spite of the 
introduction of the Arabic (or rather Hindu) system, 
"figurate numbers/' as they were called, survived the 
Middle Ages, and the term is still used, though in a more 
restricted sense. It is not a little remarkable that the 
English language has retained the name " figures," though 
it is now applied to the "Arabic" notation. 1 In other 
languages the Arabic sifr has been adopted. 

36. This way of representing numbers by " figures" 
would naturally lead up to problems of a geometrical 
nature. The dots which stood for the units were regu- 
larly called cc terms" (#/>oi, termini, "boundary stones"), 
and the spaces marked out by them were called c< fields " 
(X&pai). The question would naturally arise, "How many 
terms are required to mark out a square which is double 
of a given square ? " There is no reason for doubting 
that Pythagoras discovered that the square of the hypo- 
tenuse was equal to the squares on the other two sides ; 
but we know that he did not prove this in the same way 
as Euclid did later (I. 47). It is probable that his proof 
was arithmetical rather than geometrical ; and, as he was 
acquainted with the 3:4:5 triangle, which is always a 
right-angled triangle, he may have started from the fact 
that 3 2 + 4 2 = 5 2 . He must, however, have discovered also 
that this proof broke down in the case of the most perfect 
triangle of all, the isosceles right-angled triangle, seeing 
that the relation between its hypotenuse and its sides 
cannot be expressed by any numerical ratio. The side of 
the square is incommensurable with the diagonal. That 
is just the same sort of difficulty we meet with when we 
attempt to divide the tone or the octave into two equal 

1 The following quotations from the Netv English Dictionary are 
of interest in this connexion : 1551 RECORDS Pathw* Knowl. , . . 
" Formes (sc. produced by arrangements of points in rows) . , , whiche 
I omitte . . . considering that their knowledge appertained more to 
Arithmetike figurall than to Geometric," 16H T. Bedwell, Nat. Geom. 
Numbers, i. I, "A rationall figurate number is a number that is made 
by the multiplication of numbers between themselves." 


parts. There is no indication that Pythagoras formed any 
theory on the subject. He probably referred it simply to 
the nature of the Unlimited. 

37. Another problem which must have exercised him 
was the construction of the sphere. This he seems to 
have approached from the consideration of the dodeca- 
hedron, which, of all the regular solids, approaches most 
nearly to the sphere. Now the side of the dodecahedron 
is the regular pentagon ; and for its construction it is 
necessary to divide a line in extreme and mean ratio, the 
so-called u golden section" (Euclid, II. n). That intro- 
duces us to another " irrational magnitude/* 1 and we have 
evidence that this too played an important part as one of 
the Pythagorean mysteries. The pentalpha (so-called from 
its shape) or pentagram was used in its construction, and 

the Pythagoreans are said to have appended it to their 
letters. It continued to be used long afterwards for 
magical purposes, and we meet with it in Goethe's Faust, 
and elsewhere. Tradition represented Hippasos as the 
man who divulged Pythagorean secrets, and one story 
says he was drowned at sea for revealing the incommen- 
surability of the side and the diagonal, another that he met 
with the same fate for publishing the construction of the 

1 In the scholium on Euclid, II. 1 1 (vol. v. p. 249, Heiberg) we have 
what appears to be a Pythagorean way of expressing this. This problem, 
re told, ov &IKVVTCU 8ia j/^<a>v, "is not to be exhibited by 


regular dodecahedron. This is one of the cases where 
tradition has preserved the memory of something which 
was real and important. 

38. It was natural for Pythagoras to apply his discovery 
to the heavenly bodies, and it is extremely probable that 
he regarded the intervals between the three wheels of 
Anaximander as corresponding to the fourth, the fifth, 
and the octave. That would be the most natural explana- 
tion of the doctrine generally known by the somewhat 
misleading name of "the harmony of the spheres." There 
is no reason to believe that the celestial spheres are older 
than Eudoxos, and everything points to the conclusion 
that the Pythagoreans retained the rings or wheels of 
Anaximander. They appear in the Second Part of the 
poem of Parmenides and also in the myth of Er in Plato's 
Republic. We must further remember that there is no 
question of " harmony " in our sense of the word, but 
only of the concordant intervals, which seemed to express 
the law of the world. They yield the conception of 
" form" as correlative to " matter/' and the form is always 
in some sense a Mean. That is the central doctrine of 
all Greek philosophy to the very end, and it is not too 
much to say that it is henceforth dominated by the idea of 
or the tuning of a string. 




39. It is above all in dealing with Herakleitos that we 
are made to feel the importance of personality in shaping 
systems of philosophy. The very style of his fragments 1 
is something unique in Greek literature, and won for him 
in later times the epithet of "the dark" (6 cncorea/o'?). He 
is quite conscious himself that he writes an oracular style, 
and he justifies it by the example of the Sibyl (fr. 12) and 
of the God at Delphoi (fr. n), who " neither utters nor 
hides his meaning, but signifies it." Here we see the 
influence of what has been called the prophetic movement 
of the sixth century B.C., though we are not entitled to 
assume without more ado that Herakleitos was influenced 
by that in other respects. The truth is that his central 
thought is quite simple, and that it is still quite possible to 
disentangle it from its enigmatic surroundings. Only, 
when we have done this, we must not suppose we have 
given a complete account of the man. He is much too 
big for our formulas. 

The date of Herakleitos is roughly fixed by his refer- 
ence in the past tense to Hekataios, Pythagoras, and 
Xenophanes (fr. 16), and by the fact that Parmenides 
appears to allude to him in turn (fr. 6). This means that 
he wrote early in the fifth century B.C. He was an 

1 For references to authorities and a translation of the fragments, see 
. Gr. Ph? 63 sqq. The fragments are quoted by Bywater's numbers. 


Ephesian noble, and it appears that the ancient dignity of 
Basileus (at this date no doubt a religious office) was 
hereditary in his family ; for we are told that he resigned 
it in favour of his brother. We get a glimpse of his 
political attitude in the quotation (fr. 1 14) where he says : 
" The Ephesians would do well to hang themselves, every 
grown man of them, and leave the city to beardless lads ; 
for they have cast out Hermodoros, the best man among 
them, saying, * We will have none that is best among us ; 
if there be any such, let him be so elsewhere and among 
others.* " There can be no doubt that Herakleitos was a 
convinced aristocrat and had a sovereign contempt for the 
mass of mankind. 

But it was not only the common run of men that 
Herakleitos despised ; he had not even a good word for 
any of his predecessors. He agrees, of course, with 
Xenophanes about Homer (with whom he classes Archi- 
lochos), but Xenophanes himself falls under an equal 
condemnation. In a remarkable fragment (fr. 16) he 
mentions him along with Hesiod, Pythagoras, and Heka- 
taios as an instance of the truth that much learning 
*(7ro\vfjLa6ty does not teach men to think (voov ov Sido-Ki). 
The researches (ivropin) of Pythagoras, by which we are 
to understand in the first place his harmonic and arith- 
metical discoveries, are rejected with special emphasis 
(fr. 17). Wisdom is not a knowledge of many things ; 
at is the clear knowledge of one thing only, and this 
Herakleitos describes, in true prophetic style, as his Word 
(Xo'yo?), ^hich is "true evermore," though men cannot 
understand it even when it is told to them (fr. 2). We 
must endeavour, then, to discover, if we can, what 
, Herakleitos meant by his Word, the thing he felt he 
had been born to say, whether anyone would listen to him 
or not. 

40. In the first place, it is plain that the Word must 
be something more than the doctrine of Fire as the 
primary substance, or even the theory of Flux (iravra pet). 
If Herakleitos had merely substituted fire for the "air " of 

SOUL 59 

Anaximenes, that would only have been a further advance 
on the lines of Anaximenes himself, who had substituted 
cc air " for the water of Thales. It is not at once obvious 
either that the doctrine of flux is an improvement on that 
of rarefaction and condensation ; and, even if it were, 
such an improvement would hardly account for the tone 
in which Herakleitos speaks of his Word. It is not in 
this direction we must seek for his innermost thought. 
The doctrine of flux is, no doubt, a great scientific 
generalisation, but no single scientific discovery is 
attributed to Herakleitos. That is significant. Further, 
everything ^e are told about his cosmology shows it to 
have been even more reactionary than that of Xenophanes 
or the school of Anaximenes. On the other hand, though 
he uses the language of the mysteries, he condemns them 
in the strongest terms. The " Night-walkers, magicians, 
Bakchoi, Lenai, and Mystai" of whom he speaks (fr. 124) 
jnust be the contemporary Orphics, and we are told by 
Clement of Alexandria, who quotes the words, that 
Herakleitos threatened them with the wrath to come. 

Yet Herakleitos has one thing in common with the 
religious teachers of his time, and that is his insistence on 
the idea of Soul (^X 7 ?)* To him, as to them, the soul 
was no longer a feeble ghost or shade, but tne most real 
uthing of all, and its most important attribute was thought 
(yj/aVx,??) or wisdom (TO o-o<jt>oi>). Now Anaximenes had 
already illustrated the doctrine of "air" by the remark 
that it is breath which keeps us in life ( 9), and we 
have seen how the same idea affected the Pythagorean 
cosmology ( 28). The Delphic precept "Know thyself" 
was a household word in those days, and Herakleitos says 
c< I sought myself" (eSL^vjcrdfjajv e/xeeouroy, fr. 80). He also 
said (fr. 71) : "You cannot find out the boundaries of 
soul ; so deep a measure hath it." If we follow up these 
hints we may perhaps find ourselves on the right track. 

41. A glance at the fragments will show that the 
thought of Herakleitos was dominated by the opposition 
of sleeping and waking, life and death, and that this 


seemed to him the key to the traditional Milesian problem 
of the opposites, hot and cold, wet and dry. More pre- 
cisely, Life, Sleep, Death correspond to Fire, Water, 
Earth, and the latter are to be understood from the 
former. Now we see that the soul is only fully alive 
when it is awake, and that sleep is really a stage between 
life and death. Sleep and death are due to the advance of 
moisture, as is shown by the phenomenon of drunkenness 
(fr. 73). "It is death to souls to become water'' (fr. 68). 

. Waking and life are due to the advance of warmth and 
fire, and "the dry soul is the wisest and the best" (fr. 74). 
We see further .that there is a regular alternation of the 
two processes ; sleep alternates with waking, and life with 

1 death. Fire is fed by the exhalations of water, and these 
exhalations are in turn produced by the warmth of the 
fire. If there were no water, there could be no fire ; and, 
if there were no fire, there could be no exhalations from 
the water. 

If we look next at the macrocosm, we shall see the 
explanation is the same. Night and day, summer and 
winter, alternate in the same way as sleep and waking, 
life and death, and here too it is clear that the explanation 
is to be found in the successive advance of the wet and the 
dry, the cold and the hot. It follows that it is wrong to 
make the primary substance an intermediate state like 
"air." It must be the most living thing in the world, 

t and therefore it must be fire like the life of the soul ; and 
as the fiery soul is the wisest, so will the wisdom which 
" steers " the world be fire. Pure fire is to be seen best 
in the sun, which is lit up afresh every morning, and put 
out at night. It and the other heavenly bodies are just 
masses of pure fire ignited in a sort of basin in which they 
traverse the heavens, and this fire is kept up by exhala- 
tions from the earth. The phases of the moon and 
eclipses are due to a partial or total turning round of the 
basins. Darkness too is an exhalation from the earth of 

another kind. These last remarks prove we are not dealing 
with a scientific man, as science was understood in Italy. 


42. But, if fire is the primary form of reality, it 
seems that we may gain a clearer view of what Anaxi- 
mander had described as " separating out " ( 7), and 
Anaximenes had explained by rarefaction and condensa- 
tion " (9). The process of combustion is the key 
both to human life and to that of the world. It is a pro- 
cess that never rests ; for a flame has always to be fed by 
fresh exhalations as fuel, and it is always turning into 
vapour or smoke. The steadiness of the flame depends 
on the " measures " of fuel kindled and the " measures " 
of fire extinguished in smoke remaining constant. Now 
the world is "an everliving fire" (fr. 20), and therefore 
there will be an unceasing process of " flux/' That will 
apply to the world at large and also to the soul of man. 
"You cannot step twice into the same river " (fr. 41), 
and it is just as true that "we are and are not" at any 
given moment. " The way up and the way down," 
which are " one and the same " (fr. 69) are also the same 
for the microcosm and the macrocosm. Fire, water, 
earth is the way down, and earth, water, fire is the way up. 
And these two ways are forever being traversed in opposite 
.directions at once, so that everything really consists of two 
parts, one part travelling up and the other travelling down. 

Now Anaximander had held ( 6) that all things must 
return to the Boundless, and s^ pay the penalty to one 
another for their injustice, and what Herakleitos regarded 
ijis his great discovery seems to attach itself to this very 
pronouncement. It is just the fact that the world is " an 
everliving fire" which secures its stability ; for the same 
"measures" of fire are always being kindled and going 
out (fr. 20). It is impossible for fire to consume its 
nourishment without at the same time giving back what it 
has consumed already. It is a process of eternal "ex- 
change" (ajuLoi/3^ like that of gold for wares and wares for 
gold (fr. 22); and "the sun will not exceed his measures; 
if he does, the Erinyes, the auxiliaries of Justice, will find 
him out" (fr. 29). ForXll this strife is really justice 
'fr. 22), not injustice, as Anaximander had supposed, and 


" War is the father of all things " (fr. 44). It is just this 
opposite tension that keeps things together, like that of 
the string in the bow and the lyre (fr. 45), and though it 
is a hidden attunement, it is better than any open one 
(fr. 47). For all his condemnation of Pythagoras, Hera- 
kleitos cannot get away from the tuned string. 

But, in spite of all this, it is possible for the "measures" 
to vary up to a certain point. We see that from the facts 
of sleeping and waking, death and life, with which we 
started, and also from the corresponding facts of night and 
day, summer and winter. These fluctuations are due to 
the processes of evaporation or exhalation (avaOujuLiaarii) and 
liquefaction (x^0 which formed the starting-point of all 
early Ionian physics. Yet these fluctuations exactly 
balance one another, so that, in the long run, the 
"measures" are not exceeded. It appears certain 
that Herakleitos inferred from this periodicity the survival 
of soul in some form or other. We see that day follows 
night and summer follows winter, and we know that 
waking follows sleep. In the same way, he seems to have 
argued, life follows death, and the soul once more begins 
its upward journey. " It is the same thing in us that is 
quick and dead, awake and asleep, young and old " 
(fr, 78). That is the game of draughts that Time plays 
everlastingly (fr. 79). 

43. Such, so far as we can make it out, is the general 
view of Herakleitos, and now we may ask for his secret, 
the one thing to know which is wisdom. It is that, as the 
apparent strife of opposites in this world is really due o 
the opposite tension which holds the world together, so in 
pure fire, which is the eternal wisdom, all these opposi- 
tions disappear in their common ground. God is "beyond 
good and bad " (fr. 57, 6 1). Therefore what we must do 
to attain wisdom is to hold fast to " the common." " The 
waking have one and the same world, but sleepers turn 
aside, each into a world of his own " (fr. 95). If we keep 
our souls dry, we shall understand that good and evil are 
one, that is, that they are only passing forms of one reality 


that transcends them both. Such was the conclusion a 
man of genius drew from the Milesian doctrine of evapora- 
tion and liquefaction. 

44. For, with all his originality, Herakleitos remains 
an Ionian. He had learnt indeed the importance of soul, 
but his fire-soul is as little personal as the breath-soul of 
Anaximenes. There are certainly fragments that seem to 
assert the immortality of the individual soul ; but, when 
we examine them, we see they cannot bear this interpreta- 
*ion. Soul is only immortal in so far as it is part of the 
everliving fire which is the life of the world. Seeing that 
the soul of every man is in constant flux like his body, 
what meaning can immortality have ? It is not only true 
that we cannot step twice into the same river, but also 
that we are not the same for two successive instants. That 
is just the side of his doctrine that struck contemporaries 
most forcibly, and Epicharmos already made fun of it by 
putting it as an argument into the mouth of a debtor who 
did not wish to pay. ' How could he be liable, seeing he 
is not the same man that contracted the debt ? And 
Herakleitos is an Ionian, too, in his theology. His 
wisdom, which is one and apart from all things, "wills 
and wills not to be called by the name of Zeus" (fr. 65). 
That is to say, it is no more what the religious conscious- 
ness means by God than the Air of Anaximenes or the 
World of Xenophanes. Herakleitos, in fact, despite his 
prophetic tone and his use of religious languages, never 
broke through the secularism and pantheism of the lonians. 
Belief in a personal God and an immortal soul was already 
being elaborated in another quarter, but did not secure a 
place in philosophy till the time of Plato. 


45. We have now to consider the criticisms directed 
against the fundamental assumptions of Ionian cosmology 
from another side. That Parmenides wrote after Hera- 
kleitos, and in conscious opposition to him, seems to be 


proved by what must surely be an express allusion in his 
poem. The words " for whom it is and is not the same 
and not the same, and all things travel in opposite direc- 
tions " (fr. 6, 8), cannot well refer to anyone else, and 
we may infer that these words were written some time 
between Marathon and Salamis. We know from the 
poem that Parmenides was a young man when he wrote 
it, for the goddess who reveals the truth to him addresses 
him as " youth," and Plato says that Parmenides came to 
Athens in his sixty-fifth year and conversed with Sokrates, 
who was then "very young." That must have been in 
the middle of the fifth century B.C., or shortly after it. Par- 
menides was a citizen of Elea, for which city he legislated, 
and he is generally represented as a disciple of Xenophanes. 
It has been pointed out, however, that there is no evidence 
for the settlement of Xenophanes at Elea ( 1 6), and the 
story that he founded the Eleatic school seems to be 
derived from a playful remark of Plato's, which would 
also prove Homer to have been a Herakleitean. 1 We have 
much more satisfactory evidence for the statement that 
Parmenides was a Pythagorean. We are told that he 

t built a shrine to the memory of his Pythagorean teacher, 
Ameinias, son of Diochaitas, and this appears to rest on 

^the testimony of the inscription in which he dedicated it. 
The authorities Strabo followed, in referring to the 
legislation of Elea, expressly called Parmenides and Zeno 
Pythagoreans, and the name of Parmenides occurs in the 
list of Pythagoreans preserved by lamblichos. 2 

46. Parmenides broke with the older Ionic tradition 
by writing in hexameter verse. It was not a happy 
thought. The Hesiodic style was doubtless appropriate 
enough for the cosmogony he described in the second 

L part of his poem, but it was wholly unsuited to the arid 
dialectic of the first. It is clear that Parmenides was no 
born poet, and we must ask what led him to take this new 

1 Plato, Soph. 242 d. See E. Gr. Ph* p. 140. 
5 For all this, see E. Gr. Ph? 84 jff. 


departure. The example of Xenophanes is hardly an 
adequate explanation ; for the poetry of Parmenides is 
as unlike that of Xenophanes as it well can be, and his 
style is rather that of Hesiod and the Orphics. Now it 
has been clearly shown x that the well-known Proem, in 
which Parmenides describes his ascent to the home of the 
goddess who is supposed to speak the remainder of the 
verses, is a reflexion of the conventional ascents into 
heaven which were almost as common as 'descents into 
hell in the apocalyptic literature of those days, and of 
which we have later imitations in the myth of Plato's 
Phaedrus and in Dante's Paradiso. But, if it was the 
influence of such an apocalypse that led Parmenides to 
write in verse, it will follow that the Proem is no mere 
external ornament to his work, but an essential part of it, 
the part, in fact, which he had most clearly conceived when 
he began to write. In that case, it is to the Proem we 
must look for the key to the whole. 

Parmenides represents himself as borne on a chariot and 
attended by the Sunmaidens who have quitted the Halls 
of Night to guide him on his journey. They pass along 
the highway till they come to the Gate of Night and Day, 
which is locked and barred. The^key is in the keeping of 
Dike (Right), the Avenger, who is persuaded to unlock it 
by the Sunmaidens. They pass in through the gate and 
are now, of course, in the realms of Day. The goal of 
the journey is the palace of a goddess who welcomes Par- 
menides and instructs him in the two ways, that of Truth 
and 'Che deceptive way of Belief, in which is no truth at 
all. All this is described without inspiration and in a 
purely conventional manner, so it must be interpreted by 
the canons of the apocalyptic style. It is clearly meant to 
indicate that Parmenides had been converted, that he had 
passed from error (night) to truth (day), and the Two 
Ways must represent his former error and the truth which 
is now revealed to him. We have seen reason to believe 
that Parmenides was originally a Pythagorean, and there 
1 Diels, Parmenides Lehrgedicht, pp. 1 1 sq<? . 


are many things which suggest that the Way of Belief 
is an account of Pythagorean cosmology. In any case, it 
is surely impossible to regard it as anything else than a 
description of some error. The goddess says so in words 
that cannot be explained away. Further, this erroneous 
belief is not the ordinary man's view of the world, but an 
elaborate system, which seems to be a natural develop- 
ment of the Ionian cosmology on certain lines, and there 
is no other system but the Pythagorean that fulfils this 

To this it has been objected that Parmenides would not 
have taken the trouble to expound in detail a system he 
had altogether rejected, but that is to mistake the character 
of the apocalyptic convention. It is not Parmenides, but 
the goddess, that expounds the system, and it is for this 
reason that the beliefs described are said to be those of 
" mortals." Now a description of the ascent of the soul 
would be quite incomplete without a picture of the region 
from which it had escaped. The goddess must reveal the 
two ways at the parting of which Parmenides stands, and 
bid him choose the better. That itself is a Pythagorean 
idea. It was symbolised by the letter Y, and can be traced 
right down to Christian times. The machinery of the 
Proem consists, therefore, of two well-known apocalyptic 
devices, the Ascent into Heaven, and the Parting of the 
Ways, and it follows that, for Parmenides himself, his 
conversion from Pythagoreanism to Truth was the central 
tthing in his poem, and it is from that point of view we 
must try to understand him. It is probable too that, if 
the Pythagoreans had not been a religious society as well 
as a scientific school, he would have been content to say 
what he had to say in prose. As it was, his secession 
from the school was also a heresy, and had, like all 
heresies, to be justified in the language of religion. / 

47. All the lonians had taken for granted, that the 
primary substance could assume different forms, 1 such as 
earth, water, and fire, a view suggested by the observed 
phenomena of freezing, evaporation, and the like. Anaxi 


menes had further explained these transformations as due 
to rarefaction and condensation ( 9). That, of course, 
really implies that the stiyicture of the primary substance 
is corpuscular, and that there are interstices of some kind 
between its particles. It is improbable that Anaximenes 
realised this consequence of his doctrine. Even now it is 
not immediately obvious to the untrained mind. The 
problem was raised at once, however, by the use the 
Pythagoreans had made of the theory. According to 
them, as we have seen ( 28), the world inhaled "air," 
or void, from the boundless mass outside it, and this 
accounted for the extension of the bodies whose limits 
were marked out by the " figures." When the thing was 
put in this way, further questions were inevitable. 

48. Now the rise of mathematics in this same Pytha- 
gorean school had revealed for the first time the power of 
thought. To the mathematician of all men if is the same 
thing that ckn be thought (ecrrt voelv] and that can be 
(eo-nv efz/cu), 1 and this is the principle from which Par- 
menides starts. It is impossible to think what is not, and 
it is impossible for what cannot be thought to be. The 
great question, Is it or is it not? is therefore equivalent to 
the question, Can it be thought or not? 

Parmenides goes on to consider in the light of this 
principle the consequences of saying that anything is. In 
the first place, it cannot have come into being. If it had, 
Jt must have arisen from nothing or from something. It 
cannot have arisen from nothing ; for there is no nothing. 
It cannot have arisen from something ; for there is nothing 
else than what is. Nor can anything else besides itself 
come into being ; for there can be no empty space in 

ir rhis is how Zeller (Phil. d. griech I. 5 p. 558, . i) took fr. 5 rb 
yap a-urb voetv e<rrtv re *at ccvat, and it still seems to me the only 
possible rendering. I cannot separate elcrl vo^crcu in fr. 4, which 
2veryonMtakes to mean "are thinkable" from cart voetV in fr. 5. Nor 
Io I believe that the infinitive is ever the subject of a sentence even in 
tuch places as //. x. 174 (see Leafs note). The traditional view (given e.g. 
by Goodwin, M.T. 745) implies that iroiciv w the^ subject^ in St/ccuov 

- rovro iroidv, which is refuted by Siicaias eifu TOVTO TTOMV. 


which it could do so. Is it or is it not? If it is, then it is 
now, all at once. In this way Parmenides refutes all 
accounts of the origin of the world. Ex nihilo nihiljit. 

Further, if it is, it simply is, and it cannot be more or 
less. There is, therefore, as much of it in one place as in 
another, (That makes rarefaction and condensation im- 
possible.) It is continuous and indivisible; for there is 
nothing but itself which could prevent its/parts being in 
contact with one another. It is therefore full, a continuous 
indivisible plenum. (That is directed against the Pytha- 
gorean theory of a discontinuous reality.) 

Further, it is immoveable. If it moved, it must move 
into empty space, and empty space is -nothing, and there is 
no nothing. Also it is finite and spherical ; for it cannot 
be in one direction any more than in another, and the 
sphere is the only figure of which this can be said. 

What is (TO eov) is, therefore a finite, spherical, motion- 
less, continuous plenum, and there is nothing beyond it. 
Coming into being and ceasing to be are mere " names," 
and so is motion, and still more colour and the like. They 
are not even thoughts ; for a thought must be a thought 
of something that is, and none of these can be. 

49. Such is the conclusion to which the view of the 
real as a single body inevitably leads, and there is no escape 
from it. The " matter" of our physical text-books is just 
the real (TO eoV) of Parmenides ; and, unless we can find 
room for something else than matter, we are shut up to 
his account of reality. No subsequent system could afford 
to ignore this, but of course it was impossible to acquiesce 
permanently in a doctrine like that of Parmenides, It 
deprives the world we know of all claim to existence, and 
reduces it to something which is hardly even an illusion. 
If we are to give an intelligible account of the world, we 
must certainly introduce motion again somehow. That 
can never be taken for granted any more, as it was by the 
early cosmologists ; we must attempt to explain it if we are 
to escape from the conclusions of Parmenides. 



50. It was only possible to escape from the con- 
clusions of Parmenides on two conditions. In the first 
place, the belief that all that is is one, which had been 
held by everyone since the days of Thales, must be given 
up. There was no reason why Parmenides should have 
denied motion except this. Motion in pleno is quite con- 
'ceivable, though it would not explain anything on the 
assumption of unity. If any part of the Parmenidean 
One were to move, that could only mean that its place 
was taken at once by an equal part of it. As, however, 
this part would be precisely the same as that which 
it displaced, the result of the motion would be nit y 
and it could not be distinguished from rest. We find 
accordingly that both Empedokles and Anaxagoras, whose 
systems we have now to consider, while accepting and 
insisting on the Parmenidean doctrine that the real is 
without beginning and without end, agree in maintaining 
also that there are more kinds of real than one. The world 
we know may be explained as due to the mixture and 
separation of a number of primary " elements," The word 
ekmentum is a Latin translation of the Greek a-roL^elov^ 
"letter of the alphabet," which does not occur in this 
sense till a later date, though the conception of an 
element was quite clearly formed. Empedokles called 
his elements " roots," and Anaxagoras called his " seeds," 
but they both meant something eternal and irreducible 
to anything else, and they both held the things we 


perceive with the senses to be temporary combinations 
of these. 

The second condition that must be satisfied, if the 
world is to be explained in spite of Parmenides, is that 
some account must be given of the origin or source of the 
motion which had hitherto been taken for granted as 
something inherent in the nature of body. Accordingly, 
both Empedokles and Anaxagoras postulate causes of 
motion, which the former calls Love and Strife, and the 
latter calls Mind (VOL??). What they were feeling after was 
obviously the later physical conception of force, but it is 
equally clear that they were still unable to disentangle this 
completely from that of body. They both use language 
with regard to the forces they assume which makes it 
plain that they were pictured as something corporeal, and 
this will seem quite intelligible if we remember the part 
played by "fluids" in the science of fairly recent times. 
It is to be observed further that Empedokles felt obliged 
to assume two sources of motion, like the force of attrac- 
tion and the force of repulsion, or the centripetal and 
centrifugal forces of later days, while Anaxagoras only 
required a single force which was capable of producing 
rotation. The rotatory motion itself could account for 
everything else. 

Taking these two things together, we can under- 
stand the doctrine which is common to Empedokles 
and Anaxagoras, and which they both express in almost 
exactly the same words. It is, firstly, that there is 
in reality no such thing as coming into being (ye'mn?) 
and ceasing to be (<p6opa). That has been settled by 
Parmenides* But, secondly, it is obvious that the things 
in this world do come into being and cease to be. That 
is proved by the evidence of the senses. The only way 
in which these two things can be reconciled is by regarding 
what is commonly called coming into being as mixture, 
and ceasing to be as separation. From this it follows, in 
the first place, that the real must be such as to admit 
of mixture, or, in other words, that there must be 


different kinds of real ; and, in the second place, that 
there must be a cause of mixture and separation. 


51. Empedokles was a citizen of Akragas in Sicily, 
and he played a considerable part in his native city as a 
democratic leader. 1 His date is roughly fixed for us by 
the well-attested fact that he went to Thourioi shortly 
after its foundation in 444/3 B * c * That was probably 
after his banishment from his native city. He was, 
therefore, contemporary with the meridian splendour of 
the Periklean age at Athens, and he must have met 
Herodotos and Protagoras at Thourioi. In his case we 
know for certain that he combined scientific study with 
a mystical religion of the Orphic type, but he differed 
'from Pythagoras in the direction his scientific inquiries 
took. We know that Pythagoras was first ancj foremost 
a mathematician, while Empedokles was the founder of 
the Sicilian school of medicine. That accounts for the 
physiological interest that marks his speculations. It is 
the same difference as that between Plato and Aristotle 
at a later date. 

We are not directly concerned here with the religious 
teaching of Empedokles, though we may note in passing 
his horror of bloody sacrifices, which he justified from 
the doctrine of Rebirth or transmigration. His "Purifi- 
cations" (KaOapjuLofy, of which considerable fragments 
remain, are, indeed, our oldest and best authority for 
this type of religion. They are written in hexameters, 
and so is his more strictly philosophical poem. In this 
matter he imitated Parmenides, as is proved by his some- 
times reproducing his actual words. The only difference 
is that he was a real poet, and Parmenides was not. 

52. As has been indicated, Empedokles unreservedly 
accepts the doctrine of Parmenides that "what &" is 

1 References to authorities are given in E. Gr. ?h? 97 sqq. For a 
translation of the fragments, see &. 105, 


uncreated and indestructible, and he only escapes from 
the further conclusions of the Eleatic by introducing the 
theory of elements or " roots/' Of these he assumed 
four fire, air, earth, and water, and in some respects 
this was a return to primitive views which the Milesians 
had already left behind them ( 10). In particular, it 
was reactionary to put earth on a level with the other 
three. It must be noticed, however, that Empedokles at 
the same time made an advance by co-ordinating air with 
fire and water, instead of identifying it with vapour and 
regarding it as a transitional form between the two. He 
had in fact discovered that what we call atmospheric air 
was a body, and was quite distinct from empty space on 
the one hand and from vapour or mist on the other. 
He was doubtless led to this discovery by the polemic of 
Parmenides against the existence of empty space. The 
plain man can imagine he has a direct perception of this, 
and it was necessary for Empedokles to show he was 
wrong. This he did by means of an experiment with the 
klepsydra or water-clock. He showed that air could keep 
water out of a vessel, and that the water could only enter 
as the air escaped. This important discovery outweighs 
his error in regarding air and water as elements. He 
had no means of discovering they were not. He might, 
perhaps, have got a hint of the true nature of fire 
from Herakleitos ? but here we must remember that, so 
long as the sun and stars were believed to consist of fire, 
it was not easy to discern the truth. Even Aristotle 
adopted the four elements of Empedokles, though Plato 
and his Pythagorean friends had declared that so far from 
being " letters " (<rroxe?a), they were not even syllables. 

53. Besides these four " roots," Empedokles postu- 
lated something callecr Love (<j><A/a) to explain the attrac- 
tion of different forms of matter, and of something called 
^Strife (m/co?) to account for their separation. He speaks 
of these quite distinctly as b<5aies. The way in which they 
act seems to have been suggested by the experiment with 
the klepsydra already referred to. We start with something 


like the sphere of Parmenides, in which the four elements 
are mingled in a sort of solution by Love, while Strife 
surrounds the sphere on the outside. When Strife begins 
to enter the Sphere, Love is driven towards its centre, and 
the four elements are gradually separated from one 
another. That is clearly an adaptation of the old idea of 
the world breathing. Empedokles also held, however, 
that respiration depended on the systole and diastole of 
the heart, and therefore we find that, as soon as Strife has 
penetrated to the lowest (or most central) part of the 
sphere, and Love is confined to the very middle of it, the 
reverse process begins. Love expands and Strife is driven 
outwards, passing out of the Sphere once more in propor- 
tion as Love occupies more and more of it, just as air is 
expelled from the klepsydra when water enters it. In fact, 
Love and Strife are to the world what blood and air are 
to the body. The physiological analogy naturally influ- 
enced the founder of a medical school, who had for the 
first time formulated a theory of the flux and reflux of 
blood from and to the heart. The conception of the 
attractive force as Love is also, as Empedokles says him- 
self, of physiological origin. No one had observed, he 
tells us (fr. 17, 21-26) that the very same force men know 
in their own bodies plays a part in the life of the great 
world too. He does not seem to have thought it neces- 
sary to give any mechanical explanation of the cosmic 
systole and diastole. It was just the life of the world. 

54. A world of perishable things such as we know can 
.only exist when both Love and Strife are in the world. 
There will, therefore, be two births and two passings away 
of mortal things (fr. 17, 3-5), one when Love is increasing 
and all the elements are coming together into one, the 
.other when Strife is re-entering the Sphere and the 
elements are being separated qnce more. The elements 
alone are everlasting ; the particular things we know are 
unstable compounds, which come into being as the 
elements "run through one another" in one direction or 
another. They are mortal or perishable Just because they 


have no substance (<uW) of their own ; only the " four 
roots" have that. There is, therefore, no end to their 
death and destruction (fr. 8). 1 Their birth is a mixture 
and their death is but the separation of what has been 
mixed. Nothing is imperishable but fire, air, earth and 
water, with the two forces of Love and Strife. 

We have little information as to how Empedokles ex- 
plained the constitution of particular things. He regarded 
the four elements, which could be combined in an 
indefinite number of proportions, as adequate to explain 
them all, and he referred in this connexion to the great 
variety painters can produce with only four pigments 
(fr. 23). He saw, however, that some combinations are 
possible, while others are not. Water mixes easily with 
wine, but not with oil (fr. 91). This he accounted for 
by the presence or absence of symmetry in the "passages" 
(jropoi) or "pores" of the elements which enter into the 
mixture. It is unprofitable to inquire how he reconciled 
this view with the denial of the void he had adopted 
from Parmenides. For the rest, Aristotle attaches great 
importance to his doctrine of the "ratio of mixture" 
(\6yos TW jua'ew?), which is pretty certainly an adaptation 
of the Pythagorean theory of "blending" (Kpaa-ii) in 
fixed ratios (Aoyo*). The tuned string makes itself felt 
once more. 

55. The details of the cosmology present considerable 
difficulties. We are told that, when the elements first 
separated, fire occupied the upper hemisphere and air the 
lower. That disturbed the equilibrium of the sphere and 
produced the diurnal rotation (&'wy) of the heavens. This 
rotation, in turn, keeps the earth in the centre. The idea 
was apparently that it would naturally fall into the lower 
hemisphere, but is prevented from doing so by the lower 
hemisphere constantly becoming the upper. It is clear 
that there is great confusion of thought here. Empedokles 
has reverted to the idea of an absolute up and down in 

1 1 have adopted the interpretation of these verses suggested by Love- 
joy (Philosophical Review, xviii. pp. 371 


the world, which Anaximander had discarded already, and 
he does not seem to have been consistent even in this. 
The fiery hemisphere is day, and the airy hemisphere is 
night. The sun is only the light of the fiery hemisphere 
reflected back from the earth and gathered in a sort of 
focus. We have no means of telling how Empedokles 
worked out this singular theory in detail. We can only 
say that he was primarily a physiologist, and that astro- 
nomy was not his strong point. 

And it is certainly the case that his physiology, though 
primitive enough, makes a far more favourable impression. 
We have seen the importance he attached to respiration, 
and how he connected it with the heart's action. It was 
natural, therefore, for him to regard the blood as " what 
we think with " (o> <ppovovju.ev), 1 and to make the heart the 
central sensorium. In this he departed from the theory of 
Alkmaion of Kroton, who had discovered the importance 
of the brain for sense-perception, but he adopted from him 
the explanation of the various senses by " pores " or 
passages (iropoi). Sensation was produced by "effluences" 
(a-rroppoai) fitting into these. The origin of species was 
ascribed to the increasing action of Strife. At the begin- 
ning of this world there were undifferentiated living 
masses (pvXocpveis TVTTOL), which were gradually differen- 
tiated, the fittest surviving. Empedokles also described 
how mortal beings arose in the period when Love was 
gaining the mastery, and when everything happened in 
just the opposite way to what we see in our world. In 
that case, the limbs and organs first arose in separation, 
and were then joined together at haphazard, so that 
monsters were produced, "oxen with heads of men and 
men with heads of oxen." This strange picture of a re- 
versed evolution may possibly have been suggested by the 
Egyptian monuments. 

1 Plato, Phatdo, 9 6 b. 



56. Anaxagoras of Klazomenai is said by Aristotle to 
have been older than Empedokles, but to come " after 
him in his works " (TO?? S' epyoi? i/We^oo?). 1 It is not clear 
whether this means that he wrote later than Empedokles 
or that he was inferior to him in his achievement. His 
date is quite uncertain, but we know he settled at Athens 
and enjoyed the friendship of Perikles. Plato makes 
Sokrates attribute the eloquence of Perikles to his associa- 
tion with Anaxagoras. It was no doubt this very intimacy 
that exposed Anaxagoras to the accusation for irreligion 
(ao-/3act) which was brought against him. That is usually 
said to have happened just before the Peloponnesian War, 
but we do not really know either the date of it or the 
precise nature of the charge. " It must have been some- 
thing more definite than his speculations about the sun. 
We happen to know that even Diagoras, the typical atheist 
of those days, was not tried for his opinions, but for 
offences in language against the temples and festivals, 2 
Perikles got Anaxagoras off in some way, and he retired 
to Lampsakos, where he founded a school It is a re- 
markable fact that Plato never makes Sokrates meet him, 
though he was interested in his system, and that of itself 
suggests that the accusation for irreligion took place at 
an earlier date than the one usually given. Like a true 
Ionian, Anaxagoras wrote in prose, and considerable frag- 
ments of his book remain. 

57^ Anaxagoras lays down that the Hellenes are 
wrong in speaking of coming into being (ylveo-Oat) and 
ceasing to l>e (cwrrfXXuflr0() t They ought to call these 
" commixture " (cn^Vyeor0ou) and * c decomposition " (&a- 
Kpiveo-Qai) (fr. 17). That is almost in so many words the 
doctrine of Empedokles, with which Anaxagoras certainly 
seems to have been acquainted* In any case, it is certain 

1 References to authorities are given in jf. Gr, ?/5. 2 i ao s$g< 

2 See the speech against Andokides preserved among the works of 
Lysias (6. 17). 

"SEEDS" 77 

that he started, like Empedokles, from the Parmenidean 
account of " what is" On the other hand, Anaxagoras was 
an Ionian. We are told that he had been an adherent of 
" the philosophy of Anaximenes," and it is evident from 
the details of his cosmology that the statement is correct. 
We shall be prepared to find, then, that he started from 
quite different presuppositions, though these were also 
derived from medical sources. Medicine was the great 
-interest of the time. 

Like Empedokles, Anaxagoras postulated a plurality of 
independent elements which he called "seeds." They 
were not, however, the " four roots," fire, air, earth, and 
water ; on the contrary, these were compounds. Empe- 
dokles had supposed that bone, for instance, could be 
explained as a compound of the elements in a certain 
proportion, but this did not satisfy Anaxagoras. He 
pointed out that from bread and water arose hair, veins, 
"arteries," 1 flesh, muscles, bones, and all the rest, and he 
asked " How can hair be made of what is not hair, and 
flesh of what is not flesh ?" (fr. 10). These words certainly 
read like a direct criticism of Empedokles. 

This way of speaking, however, led to a serious mis- 
understanding of the theory. In Aristotle's biological 
works the various " tissues," some of which Anaxagoras 
enumerates, are called "homoeomerous" (oftoiopepftyjA term 
which means that all their parts are similar to the whole. 
The parts of bone are boner, and the parts of blood^ are 
blood. That is just the distinction between such things 
as bone, flesh, and blood, and "organs" like the heart or 
the lungs. There is no evidence that Anaxagoras himself 
used this terminology, and indeed it is incredible that no 
fragment containing it should have been quoted if he had. 
The Epicureans, however, attributed it to him, and they 
also understood it wrongly. They supposed it to mean that 
there must be minute particles in bread and water which 

1 The true distinction between veins and arteries was not yet known. 
The arteries were supposed to contain air and were connected with the 
wind-pipe or trachea (rpa-x^ sc. 


were like the particles of blood, flesh, and bones, and the 
adoption of this interpretation by Lucretius has given it 

58. We have seen that Anaxagoras had been an 
adherent of " the philosophy of Anaximencs," and he 
kept as close to it as he could in the details of his cos- 
mology. He could not say that everything was " air " 
more or less rarefied or condensed, for that view had 
been destroyed by Parmenides. If the world was to be 
explained at all, an original plurality must be admitted. 
He therefore substituted for the primary " air " a state of 
the world in which " all things (xp/yuara) were together, 
infinite both in quantity and in smallness" (fr. i). This 
is explained to mean that the original mass was infinitely 
divisible, but that, however far division was carried, every 
part of it would still contain all " things " (xp/Mara), and 
would in that respect be just like the whole. That is the 
very opposite of the doctrine of " elements," which seems 
to be expressly denied by the dictum that " the things 
that are in one world are not separated from one another 
or cut off with a hatchet " (fr, 8). Everything has cc por- 
tions " (uLoipai) of everything else in it. 

But if that were all, we should be no nearer an explana- 
tion of the world than before ; for there would be nothing to 
distinguish one "seed " from another. The answer to this 
is that, though each has a " portion " of everything in it, 
however minutely it may be divided, some have more of 
one thing and others more of another. This was to be 
seen already in the original undifferentiated mass where 
'all things were together"; for there the portions of air 
and "aether" (by which word Anaxagoras means fire) 
were far more numerous than the others, and therefore 
the whole had the appearance of air and " aether." Anaxa- 
goras could not say it actually was air, as Anaximenes had 
done, because he had discovered for himself or learned from 
Empedokles the separate corporeal existence of atmospheric 
air. We have some references to the experiments by which 
he demonstrated this. He used inflated skins for the 



purpose. The effort to depart as little as possible from 
the doctrine of Anaximenes is nevertheless apparent. 

59. We see, then, that the differences which exist in 
the world as we know it are to be explained by the varying 
proportions in which the portions are mingled. " Every- 
thing is called that of which it has most in it," though, 
as a matter of fact, it has everything in it. Snow, for 
instance, is black as well as white, 1 but we call it white 
because the white so far exceeds the black. As was natural, 
the " things " Anaxagoras chiefly thought of as contained 
in each "seed" were the traditional opposites, hot and cold, 
wet and dry, and so forth. It is of these he is expressly 
speaking when he says that " the things in one world are 
not cut off from one another with a hatchet" (fr. 8). 
Empedokles had made each of these four opposites a 
"root" by itself; each of the "seeds" of Anaxagoras 
contains them all. In this way he thought he could 
explain nutrition and growth ; for it is clear that the 
product of a number of " seeds " might present quite a 
different proportion of the opposites than any one of them 
if they were taken severally. 

60. The other problem, that of the source of motion, 
still remains. How are we to pass from the state of the 
world when all things were together to the manifold reality 
we know ? Like Empedokles, Anaxagoras looked to the 
microcosm for a suggestion as to the source of motion, 
but he found one such source sufficient for his purpose. 
He called it Mind (i/oi/s) ; for that is the source of motion 
as well as of knowledge in us. He did not, however, 
succeed in forming the conception of an incorporeal force 
any more than Empedokles had done. For him, too, the 
cause of motion is a sort of cc fluid." It is cc the thinnest 
of all things " (fr. 12), and, above all, it is " unmixed," that 
is to say, it has no portions of other things in it, and this 
is what gives it the " mastery," that is, the power both of 
knowing and of moving other things. Further, it enters 
into some things and not into others, and that explains the 

1 Sextus, Pyrrh. hypoU I. 33. 


distinction between the animate and the inanimate. The 
way in which it separates and orders things is by producing 
a rotatory motion (Tre^co^a-:?), which begins at the centre 
and spreads further and further. That is really all Anaxa- 
goras had to say about it, and in the Phaedo Plato makes 
Sokrates complain that he made Mind a mere deus ex 
machina (98 b). Like a true Ionian he tried to give a 
mechanical explanation of everything he could, and, when 
once he had got the rotatory motion started, he could leave 
that to order the rest of the world. 

6 1. It is hard to believe, however, that Anaxagoras 
was wholly ignorant of Pythagorean science. Oinopides of 
Chios was introducing a more highly developed geometry 
into Ionia from the west, and Anaxagoras himself is 
credited with certain mathematical discoveries. He also 
knew, though he certainly did not discover, that the 
sun is eclipsed by the interposition of the moon, and that 
the moon shines by light reflected from the sun, but he 
cannot have been able to give the true account of lunar 
eclipses, seeing that he was either ignorant of or deliberately 
rejected the discovery that the earth was a sphere. In 
this respect, too, he adhered to the doctrine of Anaximenes 
and regarded it as a disc. That being so, he had to assume 
dark bodies invisible to us to account for eclipses of the 
moon. That is probably connected with the theory which 
seems to have struck his contemporaries most. His 
attention had been directed in some way to the huge 
meteoric stone which fell into the Aigospotamos in 468/7 
B.C., and this suggested to him that portions of the earth 
might be detached and flung to a distance as from a sling 
by the rotatory motion. That had once been far more 
rapid than it is now, and so the sun, which was a mass of 
red-hot iron "larger than the Peloponnesos," and the 
moon, which was made of earth, had reached their present 
places. All this seems retrograde enough when we com- 
pare it with Pythagorean science. That was a thing the 
lonians could never really assimilate. Even Demokritos 
was nearly as backward in these matters as Anaxagoras, 


and Aristotle himself could not grasp the Pythagorean 
conception completely. 

62. Though Empedokles had distinguished Love and 
Strife as the causes of mixture and separation from the 
four elements which are mixed and separated, he continued 
to call them all " gods'* in the sense with which we are 
now familiar, and he gave the name also to the Sphere in 
which they were all mixed together, Anaxagoras seems to 
have taken the step of calling only the source of motion 
"god." In that sense and to that extent it is not incor- 
rect to call him the founder of theism. On the other 
hand, it seems to have been precisely for this that his con- 
temporaries called him an atheist. In his desire to exalt 
Nous, he seems to have followed the lead of Xenophanes 
in denying the divinity of everything else, and his state- 
ments about the sun and the moon are usually mentioned 
in connexion with the charge of irreligion brought against 
him, though we cannot tell now what that referred to, or 
whether the charge was well founded or not. We can 
only say that Perikles shared the secular spirit of the 
lonians, and it is quite conceivable that his immediate 
circle may have offended the religious susceptibilities of 
old-fashioned Athenians by ridiculing ceremonies which 
were still sacred in their eyes. 1 

1 The worship of Sun and Moon was no part of Athenian religion, 
but Anaxagoras may have ridiculed the measures prescribed by the 
c^y^roA on the occasion of the solar eclipse of 463 B.C. That, no 
doubt, would be acre/Seta. 




63. We have seen ( 46) how Eleaticism originated in 
a revolt from Pythagoreanism, and we have now to con- 
sider its detailed criticism of that doctrine. The great 
critic was Zeno. According to Plato, 1 his work, written 
when he was a young man, was intended to support the 
teaching of Parmenides by showing that the hypothesis of 
his opponents, " if things are a many >J (et xoXAa GVTL) led 
up, if thoroughly worked out, to consequences at least as 
paradoxical as his master's. We learn further from Plato 
that Zeno was twenty-five years younger than Parmenides, 
and that he was forty years old when he accompanied him 
on his celebrated visit to Athens just after the middle of 
the fifth century B.a All that agrees admirably with the 
well-authenticated statement that Perikles "heard" Zeno 
as well as Anaxagoras, and also with the accounts which 
represent Zeno as engaged in controversy with Protagoras. 
He also appears to have written against Empedokles. 2 

64, It is significant that a work of Zeno's is cited by 
the title, A Reply to the Philosophers (I5ipb$ T<W <pi\ocr6<povs) ; 
for there is reason to believe that in these days " philo- 
sopher" meant Pythagorean. At any rate, it is only if we 
regard the arguments of Zeno as directed against the 

1 Farm. 128 c* 

* References to authorities are given in E Gr* Ph? 155 /^f 


assumption that things are a many, that is to say a 
" multitude of units" (jmovdSow TrX^o?), that their real 
significance can be understood. According to the Pytha- 
gorean view, geometry was simply an application of arith- 
metic, and the point only differs from the arithmetical 
unit in so far as it is a " unit having position " (/xoj/a? Oe<nv 
expva-a). From this it ought to follow, though we need 
not suppose the Pythagoreans to have said so in so many 
words, that we should be able to say how many points 
there are in a given terminated straight line, and further 
that all magnitudes must be commensurable. The Pytha- 
goreans themselves, however, had discovered at least two 
striking instances to the contrary. We have seen that 
neither the most perfect triangle, the isosceles right-angled 
triangle, nor the most perfect solid, the regular dodeca- 
hedron, can be expressed numerically; for, as we should 
put it, \/2 and \T$ are "surds." The Pythagoreans must 
have been quite well aware of these facts, though, as we 
have seen, they probably explained them by referring them 
to the nature of the " unlimited," along with such similar 
cases as the impossibility of dividing the octave and the 
tone into equal parts. 

Zeno's arguments are directed to showing that the 
" unlimited" or, as the Eleatics call it, the continuous 
(crw/exeV, lit " hanging together " ) cannot be composed of 
units however small and however many. We can always 
bisect a line, and every bisection leaves us with a line that 
can itself be bisected. We never come to a point or unit. 
It follows that, if a line is made up out of unit-points, 
there must be an infinite number of such points in any 
given terminated straight line. Now if these points have 
magnitude, every line will be of infinite length ; if they 
have no magnitude, every line will be infinitely small. 
Again, if a point has magnitude, the addition of a point to 
a line will make it longer and its subtraction will make it 
smaller ; but, if points have no magnitude, neither their 
addition nor their subtraction will make any difference to 
the line. But that of which the addition or subtraction 

84 ZENO 

makes no difference is nothing at all. It follows that, if 
number is a sum of units (and no other account of it has 
been suggested), there is an impassable gulf between the 
discrete and the continuous, between arithmetic and 
geometry. Things are not numbers. To put the thing 
in another way, geometry cannot be reduced to arithmetic 
so long as the number one is regarded as the beginning of 
the numerical series. What really corresponds to the 
point is what we call zero. 1 

65. The celebrated arguments of Zeno concerning 
motion introduce the element of time, and are directed to 
showing that it is just as little a sum of moments as a line 
is a sum of points, (i) If a thing moves from one point 
to another, it must first traverse half the distance. Before 
it can do that, it must traverse a half of the half, and so on 
adinfinitum. It must, therefore, pass through an infinite 
number of points, and that is impossible in a finite time. 
(2) Achilles can never overtake the tortoise. Before he 
comes up to the point at which the tortoise started, the 
tortoise will have got a little way on. The same thing 
repeats itself with regard to this little way, and so on ad 
infinitum. (3) The flying arrow is at rest. At any given 
moment it is in a space equal to its own length, and there- 
fore at rest. The sum of an infinite number of positions 
of rest is not a motion. (4) If we suppose three lines, 
one (A) at rest, and the other two (B, C) moving in 
opposite directions, B will pass in the same time twice the 
number of points in C that it passes in A. From the 
interpreter's point of view this last argument is the most 
important of all. If it is directed against the view that the 
line is a sum of points and time a sum of moments, it is 
a perfectly legitimate reductio ad absurdum of these views, 
otherwise it has no meaning at all, 

1 This is the ultimate explanation of the dispute between mathe- 
maticians and historians as to whether 1900 was the last year of the 
nineteenth century or the first year of the twentieth. Astronomers call 
the year preceding i A.D. the year o, while historical chronologists make 
j A.D. the year after i B.C. 


66. The arguments of Zeno are valid only on the 
assumption that the nature of number is completely ex- 
pressed by the natural series of integers, but on that 
assumption they are unanswerable, and no other view of 
number had yet been suggested. Even rational fractions 
are unknown to Greek mathematics, and what we treat as 
such are expressed as ratios of one integer to another. 1 
Still harder was it for the Greeks to regard a surd, for 
instance, as a number, and it was only in the Academy 
that an effort was made at a later date to take a larger 
view. What Zeno actually does prove is that space and 
time cannot consist of points or moments which themselves 
have magnitude, or that the elements of a continuum can- 
not be units homogeneous with the continuum constructed 
out of them. He shows, in fact, that there must be more 
points on the line, more moments in the shortest lapse of 
time, than there are members of the series of natural 
numbers, or, what comes to the same thing, that, though 
every continuum is infinitely divisible, infinite divisibility 
is not an adequate criterion of continuity. 2 That, how- 
ever, is all he undertook to prove. We know from Plato 
that his work was an argumentum ad homines, and as such 
it is entirely successful. 


67. It is very significant that the next representative 
of the Eleatic doctrine is a Samian. As a result of the 
Persian wars, the Italic and Ionic philosophies had come 
into contact once more, and their common meeting-ground 
was Athens. Both Empedokles and Anaxagoras came 
under the influence of Parmenides, who had himself visited 
Athens along with Zeno, who apparently continued to 
reside there for some time. Anaxagoras lived at Athens 
for many years, and Empedokles took part in the Athenian 

1 Cf. e.g. the r^idXios Aoyos 3 : 2 and the tTrirpiros Xoyos 4 : 3. . 

2 1 take this way of stating the matter from Prof. A. E. Taylor's article 
" Continuity " in Hastings' Encyclopaedia of Religion and Ethics. 


colonisation of Thourioi. None of these men were them- 
selves Athenians, but they had Athenian disciples, and 
Sokrates was already in his 'teens. 

Melissos was in command of the Samian fleet that 
fought against Perikles in 441 B.C. We know nothing 
else about him. We can only guess that he had become 
acquainted with Eleaticism at Athens, and we can see that 
the modifications he introduced into it were due to cc the 
philosophy of Anaximenes," which still survived in 

68. The main arguments of Melissos are just those 
of Parmenides, except that they are expressed in simple 
Ionic prose. His great innovation was that he regarded 
the real as infinite instead of making it a finite sphere. It 
is said that he inferred its spatial infinity from its eternity, 
and he does appear to have used language that might sug- 
gest such an argument. He had, however, a much more 
cogent reason than that. The real, he said, could only be 
limited by empty space, and there is no empty space. For 
the same reason there can be no motion and no change. 
The real was, of course, corporeal, as it was for Parmeni- 
des. The statement sometimes made that Melissos held 
it to be incorporeal is based on a misunderstanding. 1 

There can be no doubt that Melissos was looked upon 
in his own day as the most advanced representative of 
Eleaticism, and "the thesis of Melissos" is an object of 
special aversion to the writer of the Hippokratean treatise 
on The Nature ofMan, while Plato makes Sokrates couple 
his name with that of the great Parmenides himself 
(Theaet. 180 e)* From a historical point of view his 
most remarkable saying is that, if things are a many, each 
one of them would have to be such as he has shown the 
One to be. That is just the formula of Atomism, as we 
shall see, and Melissos rejected it because he denied the 
existence of empty space. In that, too, he prepared the way 
for the atomic theory by making it necessary for Leukippos 
to affirm the existence of the Void. 

1 E. Gr. W 169. 


The Later Pythagoreans. 

69. It has been said already ( 27) that the Pytha- 
goreans had a singular power of adapting their theories 
to new conditions, and it is certain that at some time 
or other they felt called upon to give an account of the 
new doctrine of elements in terms of their own system* 
It is probable that this was the work of Philolaos, who 
lived at Thebes towards the end of the fifth century B.C., 
but returned to South Italy as soon as it was safe for 
Pythagoreans to show themselves in those parts once 
more. From that time forward Taras (Tarentum) was 
the chief seat of the school, and we shall hear more of 
it when we come to consider the relations of Plato with 
Archytas. For reasons I have given elsewhere, I cannot 
regard the fragments which have come down to us under 
the name of Philolaos as authentic, but for all that they 
are old and contain some valuable hints as to the develop- 
ment of Pythagorean doctrine. 1 

70. The most remarkable feature of later Pytha- 
goreanism is the way the religious side of the doctrine 
was dropped and the effort that was made to clear the 
memory of Pythagoras himself from the imputation of 
mysticism. We have the echo of this in the remains of 
Aristoxenos and Dikaiarchos, but it must be older ; for 
in their day scientific Pythagoreanism had ceased to exist. 
The statement that Hippasos of Metapontion was guilty 
of publishing a mystic discourse <c with the view of mis- 
representing Pythagoras" 2 must go back to this generation 
of the school ; for at a later date no one would have any 
interest in making it. A book by Hippasos almost cer- 
tainly existed ; for Aristotle is able to state that he made 
fire the first principle like Herakleitos. That agrees very 
well with what we can infer as to the earliest Pythagorean 
cosmology. There are all sorts of stories about this 

IE. Gr.Py5. 2 i 3 8/^. 

2 Diog. viii. 7 rbv Se MtwmKbv Aoyov 'iTnrourov . . . ?vat ycy/xxu- 
^evov ri SiapoXfj ILv6a.y6pov. 


Hippasos, who is said to have been drowned at sea or 
to have been expelled from the order, which then made a 
sepulchre for him as if he were dead. Finally, the story 
was put about that there had from the first been two grades 
in the order, Mathematicians and Akousmatics, or Pytha- 
goreans and Pythagorists, and Hippasos was represented 
as the leader of the lower grade. It is impossible, of 
course, for us to disentangle truth from falsehood in all 
this ; but we are, I think, entitled to infer that there was 
a real struggle between those who held to the Pythagorist 
religion and those who attached themselves exclusively to 
the scientific side of the doctrine. In the fourth century 
the Pythagorean scientific school expired and its place was 
taken by the Academy ; the Pythagorist religion, on the 
other hand, maintained its existence even later, as we know 
from the fragments of the comic poets. 

71. The distinctive feature of the later Pythagoreanism 
is its effort to assimilate the Empedoklean doctrine of the 
four " elements," and there is reason for believing that the 
name itself (o-Toix^O originated at this time. If Philolaos 
was the author of the theory, that is natural enough. The 
fragment of Menon's latrika recently discovered in a 
London medical papyrus has revealed the fact that he 
belonged to the Sicilian medical school, and that the 
theories of that school depended on the identification of 
the old " opposites," hot and cold, wet and dry, with the 
four elements of Empedokles. 1 The Pythagoreans had 
to find room for the elements in their system somehow, 
though they continued to resist the doctrine that they were 
ultimate. Plato has preserved this touch in his Timaeus 
(48 b), where he makes the Pythagorean protest that, 
so far from being "letters/* the four elements are not 
even syllables. 

The view they actually took of them was that they 
were " figures," or, in other words, that they were 

1 The hot and cold, wet and dry are spoken of as et&y in Tltpl 
tarptKTjs 15, and Philistion called the four elements iSu (E. Gr. Ph* 
p. ass, n. 2). 


made up of particles which had the shapes of the regular 
solids. We need not doubt that the derivation of those 
figures from the elementary triangles given in Plato's 
Timaeus is in substance Pythagorean, though, as the 
doctrine of the five regular solids was only completed by 
Theaitetos, some of the constructions must belong to a 
later date than Philolaos. 

72. The later Pythagoreans appear to have said that 
things were like numbers rather than that they actually 
were numbers, and here we shall probably be right in 
tracing the effect of Zeno's criticism. Aristotle quotes 
the doctrine in both forms, and he hardly seems to be 
conscious of any great difference between them. Further, 
he treats what is usually called the Platonic " theory of 
ideas " as practically identical with some form of Pytha- 
goreanism. That raises questions we shall have to deal 
with later ; for the present, it will be enough to consider 
what the later Pythagoreans probably meant by saying 
things were " like numbers " instead of saying that they 
actually were numbers. So far as we can see, it must 
have been something like this. For the construction of 
the elements we require, not merely groups of "units 
having position," but plane surfaces limited by lines and 
capable in turn of forming the limits of solids. Now Zeno 
had shown that lines cannot be built up out of points or 
units, and therefore the elementary triangles out of which 
the "figures" are constructed cannot be identical with 
triangular numbers such as the tetraktys. In particular, 
the isosceles right-angled triangle is of fundamental im- 
portance in the construction of the regular solids, and it 
cannot be represented by any arrangement of <c pebbles " 
(^(Jww), 1 seeing that its hypotenuse is incommensurable 
with its other two sides. It only remains for us to say, 
then, that the triangles of which the elements are ultimately 
composed are "likenesses" or " imitations" of the tri- 
angular numbers. The fateful doctrine of two worlds, 
the world of thought and the world of sense, in fact 


originated from the apparent impossibility of reconciling 
the nature of number with continuity (TO owej^) as the 
Eleatics called it, or the unlimited (TO aireipov} as the 
Pythagoreans said. There was something in the latter 
that seemed to resist the power of thought, and it was 
inferred that it could not have true reality (ova-la), but was 
at best a process of becoming (yevea-ti). You may go on 
bisecting the side and the diagonal of a square as long as 
you please, but you never come to a common measure, 
though you are always getting nearer to it. 

73. The "figures" (ei'<V) are now regarded, then, 
not as identical with the numbers, but as likenesses of 
them, and we shall not be surprised to find that, once the 
demand for a complete identification had been given up, 
an attempt was made to explain other things than the 
elements in this way. According to Aristotle, that is 
exactly what happened. The Pythagoreans went on to 
say that justice was a square number, and to give similar 
accounts of marriage, opportunity, and the like. They 
only gave a few such definitions, however, and Aristotle 
observes that they were based on mere superficial like- 
nesses between numbers and things. The most valuable 
piece of information he gives us is that Eurytos, a disciple 
of Philolaos, and therefore one of the last of the pure 
Pythagoreans, went on to express the nature of horse, 
man, and plant "by means of pebbles" or counters. 
Theophrastos said the same thing, and there seems to be 
no doubt that the statement rests on the authority of 
Archytas. Alexander gives, doubtless from the same 
source, an account of this extraordinary method. " Let 
us assume, for example," he says, "that 250 is the 
number which defines man, and 360 that which defines 
plant. Having laid this down, he took 250 counters, 
some green and some black, and others red, and all sorts 
of other colours, and then, smearing the wall with plaster 
and sketching on it a man and a plant, he proceeded 
to fix some of the counters in the outline of the face, 
some in that of the hands and some in that of other parts, 


and so he completed the outline of the man he had 
imaged by a number of counters equal in number to the 
units which he said defined the man." 

This precious testimony shows what the doctrine of 
"figures" was capable of becoming when it ventured 
beyond its proper sphere, and we must remember that 
Eurytos was not an early Pythagorean, but a leading 
man in the latest generation of the school. According to 
Aristotle, it was Sokrates that directed the theory into 
another channel by his study of moral (and aesthetic) 
forms, and Plato represents him in the Parmenidcs (130 c-d) 
as saying that at one time he had thought such things 
as man, fire, and the like should have forms as well, but 
that he had given up the idea of finding forms for every- 
thing from fear of falling into an ocean of nonsense 
(/3vdo$ (pfXvapla^. We now see what that means. Never- 
theless it is quite clear that Aristotle regards all this as 
the origin of what we call " the theory of ideas," and he 
even seems anxious to minimise the differences between 
the Platonic and the Pythagorean form of the theory, 
which did not, of course, in all cases assume such an 
extravagant form as Eurytos gave it. It was also the 
tradition of the Academy that the doctrine in question 
was of Pythagorean origin. Proklos was well read in the 
ancient commentaries on Plato, some of which went back 
to the early days of the Academy, and he distinctly attri- 
butes the original form of the theory to the Pythagoreans 
and its elaboration to Sokrates. His words are : " The 
Pythagoreans, too, had the doctrine of forms. Plato him- 
self shows that by calling the wise men of Italy friends 
of the forms (Soph. 248 a). But it was Sokrates above 
all that held the forms in honour and most explicitly 
postulated them." 1 We shall return to this when we 

1 Proclus in Farm. p. 149, Cousin : 07 v p*v yap teal napa rot? TLvBa- 
yopeioLS ^ Trepl rcov etStov fecopia, /ecu SyXoi /ecu avr&s ev 2o</>tcrr# 
TO) i/ tScov <f)i\ov$ TrpQcrayopeutov TOTJS IK 'IraAici CTO<OT;$, aAA,' o y 
juaAurra TT/oecr/Sewas /cat SiapprjSyv VTroOspwos rot et&y 


come to Sokrates ; for the present it is sufficient to point 
out that Proklos could hardly have spoken as he does if 
any other interpretation of the phrase "friends of the 
forms" (L>I> <j>l\oi) had been known in the Academy. 

74. To the same generation of the school belongs a 
remarkable advance in cosmology. It is probable that 
Philolaos still held the geocentric theory, for that is the 
only one of which we get a hint in the Phaedo ; but there 
can be no doubt that the Pythagoreans in Italy made 
the all-important discovery that the earth was one of 
the planets. They did not, indeed, make it go round the 
sun, but they postulated a Central Fire, round which the 
sun, moon, and planets all revolved. This Central Fire was 
invisible to us because the revolution of all the heavenly 
bodies was naturally explained on the analogy of the moon, 
which is the only heavenly body that can be properly 
observed by the naked eye. In other words, as the 
moon always presents the same face to us, it was supposed 
that the sun and the planets, including the earth, all 
turned the same face to the centre. It follows that we 
on the earth can see the Central Fire just as little as we 
can see the other side of the moon. In this system there 
was also a body called the Counter-earth (avrlxOoov), which 
is invisible to us because it is between the earth and 
the Central Fire. This body seems to have been assumed 
in order to explain eclipses of the moon. The shadow 
of the earth did not seem to account for them all, and 
another body casting a shadow was required. It will be 
seen that this implies the view that the moon shines by 
light reflected from the Central Fire, and it is not sur- 
prising that the same explanation should have been given 
of the sun's light. The whole cosmology of this period 
depends, in fact, on the extension of the observed facts 
regarding the moon to other bodies. 

75. Perhaps the most remarkable thing in thePytha* 
gorean doctrine of this generation is that the soul has 
come to be regarded as an "attunement" (appovla) of the 
body. That is the belief expounded by Simmias, the 


Theban disciple of Philolaos, in the Phaedo (86 b 
and we are also told that it was held by those Pytha- 
goreans who had settled at Phleious (88 d), from whom 
Aristoxenos adopted it at a later date. It cannot be 
denied that such a doctrine seems to follow quite naturally 
from the analogy of the tuned string ; but, on the other 
hand, nothing can be more inconsistent with the earlier 
Pythagorean view of the soul as something that existed 
before the body, and will continue to exist after it has left 
the body. This doctrine, on the contrary, makes the soul 
a mere function of the body, and leaves no room for the 
belief in immortality. It is probable, therefore, that its 
adoption is connected with the desire, which has been 
noted already, to drop the religious side of the Master's 



76. The first part of our story ends with Leukippos, 
the founder of Atomism ; for it was he that really answered 
the question of Thales. 1 We know next to nothing about 
his life, and his book appears to have been incorporated in 
the collected works of Demokritos. No writer subsequent 
to Theophrastos seems to have been able to distinguish 
his teaching from that of his more famous disciple. Indeed 
his very existence has been denied, though on wholly in- 
sufficient grounds. It is certain that Aristotle and Theo- 
phrastos both regarded him as the real author of the 
atomic theory, and it is out of the question that they 
should have been deceived in such a matter, especially as 
Theophrastos distinguished the teaching of Leukippos 
from that of Demokritos on certain points. 

Theophrastos was uncertain whether Leukippos was a 
native of Miletos or of Elea. The latter view is doubtless 
based on the statement that he had been a disciple of the 
Eleatics, and, in particular, of Zeno. We shall see that 
this is fully borne out by all we know of the origin of his 
doctrine, and we may infer with some probability that he 
was a Milesian who had come under the influence of Par- 
menides at Elea or elsewhere. It is not likely that it was 
at Athens ; for the atomic theory does not appear to have 
been well known there till the time of Aristotle. Plato, 
in particular, does not appear to allude to it, though it 
would certainly have interested him if he had known it. 
1 E. Gr. ?/5. 2 171 sqq. 


77. Aristotle, who in default of Plato is our chief 
authority on the subject of atomism, gives a perfectly clear 
and intelligible account of the way it arose. It almost 
appears as if he were anxious to give a more strictly his- 
torical statement than usual just because so little was known 
about atomism in the Academy. According to him, it 
originated in the Eleatic denial of the void, from which the 
impossibility of multiplicity and motion had been deduced 
Leukippos supposed himself to have discovered a theory 
which would avoid this consequence. He admitted that 
there could be no motion if there was no void, and he 
inferred that it was wrong to identify the void with the 
non-existent. What is not (TO w ov) in the Parmenidean 
sense is just as much as what is (TO ov). In other words, 
Leukippos was the first philosopher to affirm, with a full 
consciousness of what he was doing, the existence of empty 
space. The Pythagorean void had been more or less 
identified with "air," but the void of Leukippos was 
really a vacuum. 1 

Besides space there was body, and to this Leukippos 
ascribed all the characteristics of the Eleatic real. It was 
"full" (vacrTov), or, in other words, there was no empty 
space in it, but it was not one. The assumption of empty 
space, however, made it possible to affirm that there was 
an infinite number of such reals, invisible because of their 
small ness, but each possessing all the marks of the one 
Eleatic real, and in particular each indivisible (arov-ov) like 
it These moved in the empty space, and their combina- 
tions can give rise to the things we perceive with the senses. 
Pluralism was at least stated in a logical and coherent way. 
As we have seen ( 68), Melissos had already suggested 

1 The Aristotelian derivation of Atomism from Eleaticism has been 
contested, especially by Gomperz. It is tree, of course, that the Milesian 
Leukippos was concerned to vindicate the old Ionic cosmology, and, in 
particular, to save as much of the " philosophy of Anaximenes" as he 
could. So was Anaxagoras ( 61). That, however, has no bearing on 
the point at issue. Theophrastos stated distinctly that Leukippos had 
been a member of the school of Parmenides and Zeno. 


that, if things were a many, each one of them must be 
such as he held the One to be. He intended that for a 
reductio ad absurdum of pluralism, but Leukippos accepted 
it, and made it the foundation of his system. 

78. The nature of the original motion ascribed by 
Leukippos to the atoms has been much discussed. At a 
later date the Epicureans held that all the atoms are falling 
eternally downwards through infinite space, and this made 
it very hard for them to explain how they could come in 
contact with one another. There is no need to attribute 
this unscientific conception to the early atomists. In the 
first place they did not, as we shall see, regard weight as a 
primary property of the atoms ; and, in the second place, 
we have evidence that Demokritos said there was neither 
up or down, middle or end in the infinite void. 1 Aristotle 
criticised all this from the point of view of his own theory 
of absolute weight and lightness resulting in the "natural 
motions " of the elements upwards or downwards, as the 
case might be, and the Epicurean doctrine is probably the 
result of this criticism. Even Epicurus, however, had 
the grace to dispense with Aristotle's absolute lightness. 
We may therefore regard the original motion of the atoms 
as taking place in all directions, and we shall see that this 
alone will account for the formation of the worlds. 
Demokritos compared the motions of the atoms of the 
soul to that of the motes in the sunbeam which dart 
hither and thither in all directions even when there is no 
wind, 2 and we may fairly assume that he regarded the 
original motion of the other atoms in much the same way. 

79. The atoms are not mathematically indivisible like 
the Pythagorean monads, but they are physically indivisible 
because there is no empty space in them. Theoretically, 
then, there is no reason why an atom should not be as 
large as a world. Such an atom would be much the same 
thing as the Sphere of Parmenides, were it not for the 
empty space outside it and the plurality of worlds. As a 

1 Cic. de Fimbus, i. 17 ; Diog. Laert. be. 44. 
'Aristotle, de Anlma, 403 b, 31. 


matter of fact, however, all atoms are invisible. That 
does not mean, of course, that they are all the same size ; 
for there is room for an infinite variety of sizes below the 
limit of the minimum visibik. 

Leukippos explained the phenomenon of weight from 
the size of the atoms and their combinations, but he did 
not regard weight itself as a primary property of bodies. 
Aristotle distinctly says that none of his predecessors had 
said anything of absolute weight and lightness, but only 
of relative weight and lightness, and Epicurus was the 
first to ascribe weight to atoms. Weight for the earlier 
atomists is only a secondary phenomenon arising, in a 
manner to be explained, from excess of magnitude. 1 It 
will be observed that in this respect the early atomists 
were far more scientific than Epicurus and even than 
Aristotle. The__conception._pf 


place in scje&ceTand it is really one of the most striking 
illustrations of the true scientific instinct of the Greek 
philosophers that no one before Aristotle ever made use 
of it, while Plato expressly rejected it. 

80. The differences between groups of atoms are 
due to (i) arrangement and (2) position. It is not clear 
whether the illustration from the letters of the alphabet 
quoted by Aristotle was given by Leukippos or Demo- 
kritos, but in any case it is probably Pythagorean in 
origin, for it accounts satisfactorily for the use of the 
word a-roixelov in the sense of element, and that is found 
in Plato, who, as I believe, knew nothing of Atomism. 
However that may be, the points of resemblance between 
Pythagoreanism and Atomism were already noted^ by 
Aristotle, and he had direct knowledge on the subject, 
" Leukippos and Demokritos," he says, " virtually make 
all things numbers too and produce them from numbers." 
I do not see how this statement can have any meaning 
unless we regard the Pythagorean numbers as patterns 
or cc figurate numbers," and, in that case, it is still more 

1 There can be no question of mass ; for the Averts of all the atoms is 
identical, and each atom is a continuum. 


striking that Demokritos called the atoms " figures " or 
" forms " (ISeaC). The void is also a Pythagorean concep- 
tion, though, as we have seen, it was not formulated with 
precision before Leukippos. It is hardly, then, too much 
to say that the atoms are Pythagorean monads endowed 
with the properties of Parmenidean reality, and that 
the elements which arise from the various positions and 
arrangements of the atoms are, so far, like the Pytha- 
gorean " numbers." Such, at any rate, seems to be the 
view of Aristotle, though we should have been glad if he 
had explained himself more fully. 

8 1. The first effect of the motion of the atoms is that 
the larger atoms are retarded, not because they are "heavy," 
but because they are more exposed to impact than the 
smaller. In particular, atoms of an irregular shape become 
entangled with one another and form groups of atoms, 
which are still more exposed to impact and consequent 
retardation. The smallest and roundest atoms, on the 
other hand, preserve their original motions best, and these 
are the atoms of which fire is composed. It will be 
observed that it is simply taken for granted that an 
original motion will persist unless something acts upon 
it so as to retard it or bring it to a stop. To Aristotle 
that appeared incredible, and the truth had to be redis- 
covered and established on a firm basis by Galileo and 
Newton. It was really the assumption of all the earlier 
Greek philosophy. Before the time of Parmenides it was 
rest and not motion that required explanation, and now 
that Leukippos had discovered a way of escape from the 
conclusion of Parmenides, it was possible for him to revert 
to the older view. . 

82, In an infinite void in which an infinite number of 
atoms of countless shapes and sizes are constantly imping- 
ing upon one another in all directions, there will be an 
infinite number of places where a vortex motion is set 
up by their impact. When this happens, we have the 
beginning of a world. It is not correct to ascribe this to 
chance, as later writers do. It follows necessarily from the 


presuppositions of the system. The solitary fragment of 
Leukippos we possess is to the effect that " Naught 
happens for nothing, but all things from a ground (Xo^b?) 
and of necessity." It will be observed that the vortex 
theory is derived from that of Anaxagoras ( 60), which 
in turn was a development of the older Ionic doctrine. 
So far we see that Leukippos was a Milesian, but he has 
thought the matter out much more carefully than his pre- 
decessor. Anaxagoras had supposed that the analogy of a 
sling would apply, and that the larger or "heavier" bodies 
would, therefore, be driven to the furthest distance from 
the centre. Leukippos left weight out of account alto- 
gether, as a property which is not primitive, but only arises 
when the vortex has already been formed. He therefore 
looked rather to what happens in the case of bodies in an 
eddy of wind or water, and he saw that the larger bodies 
would tend towards the centre. 

83* The first effect of the vortex motion thus set up 
is to bring together those atoms which are alike in shape 
and size, and this is the origin of the four " elements," 
fire, air, earth, and water. This process was illustrated 
by the image of a sieve which brings the grains of millet, 
wheat and barley together. As this image is found also 
in Plato's Timaeus (52 e), it is probably of Pythagorean 
origin. Another image was that of the waves sorting the 
pebbles on a beach and heaping up long stones with long 
and round with round. In this process the finer atoms 
are forced out towards the circumference, while the 
larger tend to the centre. To understand this, we must 
remember that all the parts of the vortex come in contact 
(eTrtyava-ii) with one another, and it is in this way that the 
motion of the outermost parts is communicated to those 
within them. The larger bodies offer more resistance 
(avrepeio-is) to this communicated motion than the smaller, 
simply because they are larger and therefore more exposed 
to impacts in different directions which neutralise the vortex 
motion In this way they make their way to the centre 
where the motion is least, while the smaller bodies arr 


squeezed out towards the circumference where it is greatest 
That is the explanation of weight, which is not an "occult 
quality," but arises from purely mechanical causes. 

84. When we come to details, we find that Leukippos 
showed himself a true Ionian. His Eleatic teachers doubt- 
less warned him off the Pythagorean cosmology, but they 
could not give him a better. It was natural, then, that 
he should turn to the theories of his distinguished fellow- 
citizen Anaximenes, and the little we know of his system 
shows that he did so, just as Anaxagoras had done before 
him. He deliberately rejected the Pythagorean discovery 
that the earth was spherical, a discovery of which he 
cannot have been ignorant, and taught that it was in shape 
" like a tambourine," resting on the air. The reason why 
it sloped toward the south was that the heat there made 
the air thinner and therefore less able to support it. In 
fact, the Atomists rejected the Pythagorean theory of the 
earth exactly as Anaxagoras had done, and it was only 
the fusion of Eastern and Western cosmology at Athens 
that finally established the new view. Though Aristotle's 
earth is in the centre of the universe, it never occurs to 
him to doubt its spherical shape. 

85. It is not worth while to follow in detail the 
application of the atomic theory to particular phenomena, 
and the atomic explanation of sensation and knowledge 
will be better kept till we come to Demokritos, to whom 
it was chiefly due. All we need say further here is that 
Leukippos has answered the question of Thales in the 
sense in which Thales had asked it, and no further 
advance was possible on these lines. Before that could 
take place it was necessary that attention should be 
directed to the kindred problems of knowledge and of 
conduct, and we shall see in the next book how that came 
about. The very completeness of the mechanical theory 
of the world which had now been given brought science 
to a standstill for a time, and it also provoked a 
revolt against cosmology. On one side that came from 
specialists in the particular sciences, especially medicine, 


who disliked the sweeping generalisations of the cos- 
mologists, and maintained the right of each science to 
deal with its own province. The Hippokratean treatise 
on Ancient Medicine (by which is meant the art of 
medicine based on experience and observation, as con- 
trasted with the new-fangled medical theories of the 
school of Empedokles and others) is the best evidence 
of this. On the other side, there was a revolt against 
science which proceeded from men whose chief interest 
was in practical life. How do you know these things are 
true, they said, and even if they are, what does it matter 
to us ? Those two questions can only be dealt with by 
a theory of knowledge and a theory of conduct 




Law and Nature 

86. We have now to consider a period of breakdown 
and reconstruction. Science had done all it could to 
make the world intelligible, and the result was a view 
of reality in flat contradiction to the evidence of the 
senses. Apparently it was not this world science explained 
but another one altogether. What, then, are we to say 
about this world? Why should we regard the world 
of science as truer than it ? After all, that world is a 
product of human thinking, and how can we tell that 
thought is not as misleading as sense is said to be ? 
Science proceeds on the assumption that there is some 
fundamental reality (^wny) which we can discover, but 
what guarantee have we for that ? It is very plain that 
men's views of right and wrong, fair and foul, vary from 
people to people, and even from city to city, so there is 
no fundamental reality in them at any rate. In the same 
way the scientific schools only agree in one thing 
namely, that all other schools are wrong. It is surely 
just as unlikely that any of these schools should possess 
the truth as that any of the nations, Hellenic or barbarian, 
should have established among themselves the true law of 
nature. Such were the thoughts that must have kept 
suggesting themselves to cultivated men in the middle of 
the fifth century B.C. 

It is very significant that the difficulties which were felt 


as to knowledge and conduct should both have been 
summed up in the same antithesis, that of nature 
(<jtw<n?) and law (yo'/xo?), though the latter term has to do 
primarily with conduct and the former with knowledge. 
This shows that the two problems were felt to be the 
same. The use of the term Law was evidently due to 
the great legislative activity of the preceding centuries. 
In early days the regularity of human life had been 
far more clearly apprehended than the even course 
of nature. Man lived in a charmed circle of law and 
custom, but the world around him still seemed lawless. 
So much was this so that, when the regular course 
of nature began to be observed, no better name could 
be found for it than Right or Justice (Six*}), a word 
which properly meant the unchanging custom that 
guided human life. We have seen that Anaximander 
spoke of the encroachment of one element on another as 
" injustice " ( 6), and, according to Herakleitos, it is 
the Erinyes, the avenging handmaids of Right, that 
keep the sun from " overstepping his measures " ( 42). 
But a code of laws drawn up by a human lawgiver whose 
name was known, a Zaleukos, or a Charondas, or a Solon, 
could not be accepted in the old way as part of the 
everlasting order of things. It was clearly something 
"made," and it might just as well have been made 
otherwise or not made at all, A generation that had 
seen laws in the making could hardly help asking itself 
whether the whole of customary morality had not after 
all been made in the same way. That is why we find 
the word which is properly applied to the legislator's 
activity (0<?W) * used synonymously with law (v6vo$) in 
this connexion. 

The best evidence of this state of feeling is the work of 
Herodotos. He must certainly have known Protagoras 
at Thourioi, and some have thought that they could 
detect the influence of Protagoras in his work. It may 
be so, but it is just as likely that he is the mouthpiece of 
1 Whence " positive " as opposed to " natural " law. 


a feeling which was widely spread at the time, and to 
which Protagoras gave expression in another form. In 
any case, it is quite wrong to regard him as a representa- 
tive of old-fashioned morality and religion. He is utterly 
sceptical, and his respect for conventions is due to his 
scepticism, just like that of Protagoras. The strongest 
proof he can give of the madness of King Cambyses is 
that he laughed at the rites and customs of other nations 
as if his own were a bit less artificial " If we were to set 
before all men a choice, and bid them pick out the best 
uses (I/O/AOI) from all the uses there are, each people, after 
examining them all, would choose those of their own 
nation." So " it is not likely that any one but a madman 
would laugh at such things," and Pindar was right in 
saying that "Law is king of all." 1 

The "Sopbists" 

87. It is usual to speak of the men we have now to 
deal with as " the Sophists,'' and so they called themselves 
and were called by others. For us, however, the name 
Sophist is apt to be misleading in more ways than one. 
It is misleading if it is used to indicate a contrast between 
these men and the thinkers and teachers of an earlier 
generation. Herodotos calls Pythagoras a Sophist (iv. 95). 
It is still more misleading if it makes us think of them as 
forming in any sense a sect or school, or even as teachers 
with identical aims and methods. There is the further 
difficulty that, by the fourth century B.C., the word had 
already begun to acquire the meaning it still bears in 
ordinary language. This seems to have originated with 
Isokrates, who was anxious to keep what he called " philo- 
sophy "distinct from intellectual pursuits of another order. 
Plato, too, for reasons we shall have to consider, was 
anxious to distinguish the Sophist from the Philosopher, 

1 Herod, iii. 38. The quotation from Pindar is the more significant 
that Pindar meant something quite different (see below, 97). It was 
therefore a familiar " text " that could be made to mean anything. 


and in one of his later dialogues defines the former as 
a paid huntsman of rich and distinguished young men. 
Aristotle formulated all that, and defines the Sophist as 
one who makes money out of apparent wisdom. l 

Now we must observe that the Sophists here referred to 
are primarily contemporaries of Isokrates, Plato, and Aris- 
totle themselves, not the distinguished teachers of the fifth 
century who commonly go by the name, and we have no 
right to transfer the polemics of a later generation to that 
of Protagoras and Gorgias. Aristotle's definition of the 
Sophist must, therefore, be left out of account altogether, 
and we shall see that the people Isokrates calls Sophists 
are certainly not those the word most naturally suggests 
to a modern reader. Plato is a safe guide when he is 
dealing by name with the great Sophists of the fifth cen- 
tury ; his general discussion in the dialogue entitled The 
Sophist has, we shall see, another bearing. 

We do learn from Plato, however, that, even in the fifth 
century, there was a prejudice against the name which 
made it possible for it to acquire the unfavourable sense it 
had in the fourth. That prejudice took two forms, an 
aristocratic and a democratic. From the democratic point 
of view, indeed, there was no blame attaching to the title 
<ro(pi(rTJi$ that did not equally attach to the word o-o<pos 
itself. To be "too clever" was always an offence, and in 
the Apology it is just the charge of being a " wise man " 
that Sokrates is most eager to rebut. From the aristo- 
cratic point of view, the name was open to another 
objection. Its very form suggested professionalism, 2 a 
thing the high-born Hellene shrank from instinctively. 
Above all, the fact that these distinguished men were 
foreigners made them unpopular at Athens. The Athenian 
public was full of prejudices, and that against "the for- 
eigner " was particularly well developed. It was in part 

1 Plato, Soph. 223 b ; Arist. Soph. EL 165 a, 22. 

* The tro^wrnjs makes a profession of " being clever" or "playing the 
wit " (TO 0"o<ir#<u) just as the KiQapurrris makes a profession of playing 
on the lyre. 


the cause and in part the effect of the growing stringency 
with which the privilege of citizenship was guarded. An 
Athenian orator or comic poet had no more effective 
weapon than the charge of foreign extraction. We know 
something of such nationalism in our own day, and in 
democratic Athens it was a very potent force indeed. 
Such considerations as these explain why Plato represents 
Protagoras as wearing the name of Sophist with a certain 
bravado. 1 

This view is more or less common ground at the present 
day ; but it can hardly be said that all its consequences 
have been fully realised. German writers in particular 
continue to be much influenced by a superficial analogy 
between the "age of the Sophists" and the eighteenth 
century Aufkldrung^ with the result that the Sophists are 
represented either as subverters of religion and morality, 
or as champions of free thought, according to the personal 
predilections of the writer. The truth is rather that, 
so far as there is any parallel to the Aufkl'drung in the 
history of Greek thought at all, it occurs much earlier, 
and Xenophanes, not Protagoras, is its apostle. It is not 
to religion but to science that Protagoras and Gorgias take 
up a negative attitude, and we shall never understand them 
if we lose sight of that fundamental distinction. The "age 
of the Sophists " is, above aU, an age of reaction against 

88. It has been pointed out that the Sophists did not 
constitute a school, but it is true for all that that their 
teaching had something in common. They all aim chiefly 
at practical ends. Their profession is that they teach 
"goodness" (aperri\ and that is explained to mean the 
power of directing states and families aright. In practice 
this was apt to work out in a curious way, especially in a 
democratic state like Athens. The Sophists quite naturally 
taught people who could pay them, and these were generally 
the well born and well-to-do, who were the natural prey of 
the democracy. To a large extent, then, what they taught 

t. 317 b. 


was the art of succeeding in a democratic State when you 
do not yourself belong to the ruling democracy, and, in 
particular, the art of getting off when you are attacked in 
the courts of law. That is the questionable side of the 
Sophist's work, but it is hardly fair to make it a ground of 
accusation against the men themselves ; it was the natural 
outcome of the political conditions of Athens at the time. 
There is no reason to doubt that Protagoras was perfectly 
sincere in his profession that he was a teacher of "good- 
ness " : only the goodness demanded by his clients was 
apt to be of a rather odd kind, and in practice his teaching 
became more and more confined to the arts of rhetoric 
and disputation. He would never have been entrusted 
by Perikles with the highly responsible task of framing a 
code of laws for Thourioi unless he had really possessed 
considerable skill in politics and jurisprudence ; but the 
young men he was called on to train were more likely to 
be engaged in conspiracies against the State than in legis- 
lation. That was not his fault, and it will help us to 
understand the Sophists much better if we bear in mind 
that, from the nature of the case, they were compelled to 
depend mainly for their livelihood on the men who after- 
wards made the oligarchic revolutions. In that sense only 
were they the products of democracy ; what a sincere 
though moderate democrat really thought of them we 
may gather from what Anytos is made to say in Plato's 
Meno (91 c sqq.}> 


89. The earliest Sophist m the sense just explained 
was Protagoras ot Abdera. In the dialogue called by his 
name, Plato has described his second visit to Athens. 
He had been there once before when Hippokrates r the" 
Athenian youth who asks Sokrates for an introduction to 
him, was still a boy This time there is a great gathering 
of Sophists from all parts of the Hellenic world in the 
house of Kallias, son of Hipponikos, who was known to 
have spent more money on Sophists than any man of his 



day. It is obvious that such a gathering would have 
been impossible at any time during the first stage of the 
Peloponnesian War. Alkibiades is quite a lad, though he 
has a beard coming (309 a). Protagoras is represented 
as much older than Sokrates, and indeed he says (3170) 
there is no one in the company (which includes Hippias 
and Prodikos) whose father he might not be, and also that 
he has been engaged in his profession for many years. 
All through he addresses his hearers as men who belong 
to a younger generation. In the Hippias maior (282 e) 
Hippias is made to say that Protagoras was " far older " 
than he was. From the Meno we get further information. 
That dialogue is supposed to take place before the expedi- 
tion of Cyrus (401 B.C.) in which Meno took part, and 
Protagoras is spoken of (91 e) as having died some con- 
siderable time before, when he was seventy years old and 
had been forty years in practice, in which time he had made 
more money than Pheidias and any other ten sculptors put 
together. Lastly, in the Theaeteius, a dialogue supposed 
to take place just before the trial of Sokrates, Protagoras 
is spoken of as one long dead. 

Now all these statements are perfectly consistent with 
one another, and the total impression they make on us 
would not be affected by one or two minor anachronisms, 
if such there are. 1 They mean that Protagoras was born 
not later than 500 B.C., that his second visit to Athens 
cannot have been later than 432 B.C., and may have been 
some years earlier, and that he died in the early years of 
the Peloponnesian War. These dates are perfectly con- 
sistent with the well-attested fact that he legislated for 
Thourioi in 444/3 B.C., 2 and they are quite inconsistent 

1 Though Protagoras is represented as putting up Trapa KaAAtp, rov 
'ITTTTOVLKOV (3113), that does not imply that Hipponikos was dead. 
In the Republic (328 b) Sokrates and the rest go ts TLoXefMpxov, though 
Kephalos is certainly living. The imperfect *XP*] TO (3 I 5^) rather 
implies that Hipponikos was still living. 

2 The traditional date of Protagoras is based solely on this. Everyone 
connected with Thourioi is supposed to have " flourished " in the year 


with the statement that he was prosecuted and condemned 
for impiety in the time of the Four Hundred (411 B.C.). 
Indeed, Plato represents Sokrates as saying things which 
make it impossible to believe Protagoras was ever pro- 
secuted for impiety at all. 1 In the Meno a special point is 
made (91 e) of the fact that throughout his long life 
no one ever suggested that he had done any harm to his 
associates, and that his good name remained unsullied 
down to the supposed date of the dialogue, several years 
after his death. Further, there is no reference to any 
accusation of Protagoras in the Apology, though such a 
reference would have been almost inevitable if it had ever 
taken place. 2 Sokrates has to go back to the trial of 
Anaxagoras to find a parallel to his own case. It is there- 
fore safer to dismiss the story altogether. 

The portrait Plato has drawn of Protagoras has been 
called a caricature, but there does not seem to be much 
ground for such a view. In the first place, we must 
observe that he does not speak of him in his own person. 
It is Sokrates that describes him, and he only applies to 
Protagoras the irony he habitually applied to himself. 

of its foundation, and to " flourish " is to be forty years old. For that 
reason Empedokles, Herodotos, and Protagoras are all said to have been 
born in 484/3 B.C. It seems probable, however, that a lawgiver would 
be over forty. 

1 The statement that Protagoras was accused by Pythodoros, son of 
Polyzelos (Diog. Laert. ix. 54), sounds circumstantial, but the next 
words, " but Aristotle says it was Euathlos," shows that this notice really 
refers to the celebrated "Suit for his Fee" (Ac'*?? we/> fjuarOov). The 
story was (ib. ix. 55) that Euathlos was to pay the fee when he had won 
his first case. When Protagoras demanded it, he replied, " I have not 
won a case yet." The answer was that Protagoras would sue him, and 
then he would have to pay. " If I win, because I have won ; if you 
win, because you have won." 

* It is worth while noting that the oldest form of the story appears to 
have made the accusation of Protagoras subsequent to that of Sokrates 
(cf. Timon, fr. 5 Diels). He was supposed to be a contemporary of Plato 
owing to the common confusion of Sokrates and Plato, and was accord- 
ingly made a disciple of Demokritos, who really belonged to a later 


Such good-humoured raillery as there Is refers mainly to 
the enthusiastic admirers of the great man. Indeed, we 
are made to feel that Sokrates has a genuine respect for 
Protagoras himself. It is true that in the Theaetetus he 
does caricature his teaching, but he immediately confesses 
that it is a caricature, and goes on to give a much more 
sympathetic account of it. 

90. There is considerable uncertainty about the 
number and titles of the works of Protagoras, which is 
due, no doubt, to the fact that titles, in the modern sense, 
were unknown in the fifth century. 1 The work Plato 
refers to as The Truth ('AA$taa) is probably identical with 
that elsewhere called The Throwers (Kara/SoAAoi/re?, sc. 
\6yot), 2 and was no doubt the most important. If we 
reject the story that Protagoras was accused of impiety, we 
must also, of course, reject that of the destruction of all 
copies of his work by public authority. In any case, it is 
absurd. The book is represented as widely read long 
after Protagoras died. In the Theaetetus of Plato (152 a) 
the lad from whom the dialogue takes its name says he 
has read it often, and in the Helen (10. 2) Isokrates 
says : " Who does not know that Protagoras and the 
Sophists of that time have written elaborate works and left 
them to us ?" And even if the Athenians had been so 
silly as to burn all the copies they could find at Athens, 
there must have been many others scattered through the 
Greek world from Abdera to Sicily, and these would not 
be at the mercy of the Athenian authorities. It is clear, 
then, that the book was extant and widely read when 
Plato quoted it, and that it would have been impossible for 
him to interpret the doctrine of Protagoras in a sense not 
really suggested by it. 

1 This statement refers primarily to prose works. Dramas had titles of 
a sort (Le, they were called after the choms or the protagonist), and 
Plato followed this custom in naming his dialogues. 

2 Metaphors from wrestling are regular in this connexion, and /cara- 
fidXX,iv means " to throw/' The phrase KarafiaXXciv ra<$ awr&Jcreis 
became technical for attacks upon sensation as a source of knowledge. 



91. That doctrine is the famous one that " Man is the 
measure of all things, of things that are that they are, and 
of things that are not that they are not." The meaning 
of this dictum has been much canvassed, but the curious 
use of the word " measure" has not been sufficiently 
remarked. We have become so accustomed to the phrase 
that it hardly strikes us as peculiar, and yet it is surely not 
the most obvious way of expressing any of the meanings 
that have been attributed to Protagoras. Why "measure " ? 
To understand this, we should probably start from the 
arithmetical meaning of the word. It is recorded that 
Protagoras attacked mathematics, and in particular the 
doctrine that the tangent touches the circle at a point. 
There must, he urged, be a stretch for which the straight 
line and the circle are in contact. 1 It is probable, then, 
that his use of the word " measure" was due to the contro- 
versies about incommensurability which were so rife in 
the fifth century. The geometers tell us, he may have 
said, that the side and the diagonal of the square have no 
common measure, but in cases like that man is the 
measure, that is, they are commensurable for all practical 
purposes. Theories that set themselves in opposition to 
the commonsense of mankind may safely be ignored. We 
shall find that this is just the position Protagoras took up 
on other questions. In the great controversy about Law 
and Nature he is decidedly on the side of the former. 

In this connexion it is interesting to note that tradition 
represents Protagoras as having met Zeno at Athens, 
which he may well have done, and there was a dialogue in 
which the two men were introduced discussing a question 
closely bound up with the problem of continuity. A 
quotation from it has been preserved, and its authenticity 
is guaranteed by a reference to it in Aristotle. 2 "Tell 
me, Protagoras," said Zeno, " does a single grain of millet 

l Arist. Met. B, 2. 9983, 2. 

2 Simplicius, Phys. 1108, 1 8 (R.T, 131), Ar. Phys. 2503, 20. That 
such dialogues existed is the presupposition of Plato'? Ptrmcnides. It 
professes to be one of them, 


make a noise in falling or the ten-thousandth part of a 
grain ?" And when he said it did not, Zeno asked him, 
a Does a bushel of millet make a noise when it falls or 
not ?" And, when he said it did, Zcno replied, " What 
then ? Is there not a ratio of a bushel of millet to one 
grain and the ten-thousandth part of a grain ?" When 
he said there was, Zeno replied, " Well, then, will not the 
ratios of the sounds to one another be the same ? As the 
sounding objects are to one another, so will the sounds be to 
one another ; and, if that is so, if the bushel of millet makes 
a noise, the single grain and the ten-thousandth part of a 
grain will make a noise." This quotation proves at least 
that it was thought appropriate for Protagoras and Zeno 
to discuss questions of the kind, and so confirms the view 
that it really was the Eleatic dialectic which made men turn 
away from science. Moreover, Porphyry said he had come 
across a work of Protagoras containing arguments against 
those who introduced the doctrine that Being was one. 1 

92. But who is the " Man" who is thus " the measure 
of all things" ? Plato more than once explains the meaning 
of the doctrine to be that things are to me as they appear 
to me, and to you as they appear to you. It is possible 
that this may not be a verbal quotation, but it is hard to 
believe that Plato could have ventured on such an inter- 
pretation if there was no ground for it. It also seems to 
me that the modern view which makes Protagoras refer, 
not to the individual man, but to " Man as such," attri- 
butes to him a distinction he would not have understood, 
and would not have accepted if he had. The good faith 
of Plato is further confirmed by the hint he gives us, when 
he does go on in the Theaetetus to develop an elaborate 
sensationalist theory from the dictum of Protagoras, that 
it was not so developed by Protagoras himself. He says 
it was something he kept back from the common herd and 
only revealed to his disciples " in a mystery." We could 
hardly be told more plainly that the theory in question 
was not to be found in the book of Protagoras itself. 
s. T.E. r. 3, 25 (Bernays, Ges. Abh. I 121). 


Nor does Plato stand alone in his interpretation of this 
dictum. Demokritos, who was a younger fellow-citizen 
of Protagoras, understood it precisely in the same way. 
We learn from Plutarch that the Epicurean Kolotes had 
accused Demokritos of throwing human life into confusion 
by teaching that " nothing was such rather than such" 
(ovSev f*\\ov rolov rj Toioity. Plutarch (or rather his 
authority) replies that, so far from holding this view, 
Demokritos combated Protagoras who did hold it, and 
wrote many convincing arguments against him. 1 It is 
impossible to ignore that, and the testimony of Demo- 
kritos is not only of the highest value in itself, but is, of 
course, quite independent of Plato's. 

The practical inference to be drawn from all this is that 
on every subject it is possible to make two opposite state- 
ments (Aoycn), both of which are " true," though one may 
be " weaker " and another " stronger." It is the business 
of the disputant to make the weaker statement the stronger 
(TW TJTTW \6yov Kpeirro) TTOMIV), and that is an art which 
can be taught. It is important to notice that this is not 
in itself an immoral doctrine. Plato distinctly tells us that 
though, according to Protagoras,' all beliefs are equally 
true, one belief may nevertheless be better than another, 
and he seems to have regarded as "better" the beliefs 
which were most in accordance with those of the man in 
a normal condition of body and mind. People who have 
jaundice see all things yellow, and just so it is possible for 
a man to have his moral beliefs coloured by some abnormal 
condition of soul. The things that appear yellow to the 
jaundiced eye really are yellow to it, but that does not 
alter the fact that it would be better for the sick man if 
they appeared different to him. His belief would not be 
truer, but it would be better. In the same way, then, as 
it is the business of the doctor to bring his patient's body 
into such a condition that he may see normally, so it is the 
business of the Sophist to make the better statement, which 

J Plut. adv. Col. 1 1 08 f. jf. Cf. Scxtus Empiricus, adv. Hath, vii 

LAW 117 

may be the weaker in a given case, not only better but 

93. This explains further how it is that Plato repre- 
sents Protagoras as a convinced champion of Law against 
all attempts to return to Nature for guidance. He was a 
strong believer in organised society, and he held that 
institutions and conventions were what raised men above 
the brutes. That, at any rate, is the meaning of the 
myth Plato puts into his mouth in the dialogue called by 
his name. So far from being a revolutionary, he was the 
champion of traditional morality, not from old-fashioned 
prejudice, but from a strong belief in the value of social 
conventions. In this sense, he not only professed to teach 
" goodness H himself, but he believed it was taught by the 
laws of the state and by public opinion, though not 
perhaps so well. He had a profound belief in the value 
of such teaching, and he considered that it begins in early 
childhood. The less he could admit anything to be truer 
than anything else, the more sure he felt that we must 
cleave to what is normal and generally recognised. 

The attitude of Protagoras to religion is generally 
looked at in the light of the highly improbable story of 
his accusation for impiety. We still have a single sentence 
from his work On the Gods, and it is as follows : " With 
regard to the gods, I cannot feel sure either that they are 
or that they are not, nor what they are like in figure ; for 
there are many things that hinder sure knowledge, the 
obscurity of the subject and the shortness of human life." 
There is surely nothing impious in these words from any 
point of view, and certainly there is none from the Greek. 
Speculative opinions on subjects like these were no part of 
Greek religion, which consisted entirely in worship and 
not in theological affirmations or negations. 1 And, in any 
cas,e, the sentence quoted might just as well be the prelude 
to a recommendation to worship according to the use of 
one's native city (VOJJ.H) TroXewi) as to anything else, and 
such a recommendation would be in complete harmony 


with the other views of Protagoras. If we cannot attain 
sure knowledge about the gods by ourselves, we shall do 
well to accept the recognised worship. That is what we 
should expect the champion of Law against Nature to say. 

Hippias and Prodikos. 

94. The other Sophists mentioned as present in the 
house of Kallias arc of no great importance for the history 
of philosophy, though they are of considerable interest 
as typical figures. Hippias of Elis is chiefly memorable 
for his efforts in the direction of universality. He was 
the enemy of all specialism, and appeared at Olympia 
gorgeously attired in a costume entirely of his own making 
down to the ring on his finger. He was prepared to 
lecture to anyone on anything, from astronomy to ancient 
history. Such a man had need of a good memory, and 
we know that he invented a system of mnemonics. There 
was a more serious side to his character, however. This 
was the age when men were still sanguine of squaring the 
circle by a geometrical construction. The lunules of Hip- 
pokrates of Chios belong to it, and Hippias, the universal 
genius, could not be behindhand here. He invented the 
curve still known as the quadratrix (reTpayuvlfyvcra), 
which would solve the problem if it could be mechanically 
described. Prodikos of Keos is chiefly known nowadays 
for the somewhat jejune apologue of the Choice of 
Heraklcs which Xenophon has preserved. We shall see 
presently how important the personality of Herakles was 
at the time. The chief work of Prodikos, however, seems 
to have been the discrimination of synonyms, a business 
which may possibly have been important in the infancy 
of grammar, Protagoras too contributed something to 
grammar. He called attention to the arbitrary character 
of certain grammatical genders, no doubt in illustration of 
the reign of Law or convention, and his classification of 
sentences into command, wish, etc. prepared the way for 
the distinction of the moods. 



95. Gorgias of Leontinoi in Sicily came to Athens as 
ambassador from his native city in 427 B.C., when he was 
already advanced in years. His influence, therefore, be- 
longs to a later generation than that of Protagoras, though 
he need not have been younger than Hippias and Prodi- 
kos. He had, it seems, been a disciple of Empedokles, 
and we learn incidentally from Plato's Meno (76 c) that 
he continued to teach that philosopher's doctrine of 
" effluences " even in his later days, when he had retired 
to Larissa in Thessaly. He is said to have lived to a 
great age, but no precise date can be given for his death. 
It is evident from Plato's account of him that he was not 
so much a teacher of politics, like Protagoras, as a teacher 
of rhetoric. That is accounted for by the change in the 
political situation brought about by the Peloponnesian 
War and the death of Perikles. The relations between 
the democracy and the well-to-do classes were becoming 
more and more strained, and the importance of forensic 
rhetoric was accordingly increased. What Gorgias did 
was to introduce to Athens the methods of persuasion by 
means of artistic prose which had been elaborated during 
the struggle of classes in Sicily. His influence on Athenian 
literature, and through it on the development of European 
prose style in general, was enormous. It does not concern 
us here, except incidentally, but it is worth while to note 
that the terms " figure " (eT&>?, <r^jma) and cc trope" 
(Tpo7ro9\ which he applied to the rhetorical devices he 
taught, are apparently derived from Pythagorean musical 
theory ( 32), and mean primarily the arrangement of 
words in certain patterns. 1 

96. Like Protagoras, Gorgias had been driven by the 
Eleatic dialectic to give up all belief in science. Prota- 
goras, as we have seen, fell back on " common sense," but 
Gorgias proceeded in a much more radical fashion. If 

1 Taylor, Van* Socratica, i. p. 206, 0.1. C also the uses of eZSos and 
for poems. 


Protagoras taught that everything was true, Gorgias 
maintained there was no truth at all. In his work entitled 
On Nature or the non-existent (TLepi (pvcrew $ rod M cWo?) x 
he sought to prove (i) that there is nothing, (2) that, 
even if there is anything, we cannot know it, and (3) that, 
even if we could know it, we could not communicate our 
knowledge to anyone else. We have two apparently 
independent accounts of the arguments by which he 
established these positions ; but, though they agree 
generally with one another, they are obviously paraphrases 
in the language of a later time. We can still see, however, 
that they were borrowed in the main from Zeno and 
Melissos, and that is a mark of their being in substance 
authentic. Isokrates, who had been a disciple of Gorgias, 
mentions his assertion that Nothing is in the Helen (10.3), 
and he couples his name with those of Zeno and Melissos, 
thus confirming in a general way the later accounts. The 
reasoning of Zeno and Melissos was of a kind that is apt 
to cut both ways, and that is what Gorgias showed. The 
argument given as peculiar to himself was to this effect. 
"What is not" is not, that is to say, it is just as much as 
"what is." The difficulty here raised is one that was not 
cleared up till Plato wrote the Sophist. We shall consider 
it when we come to that. 

97. In the ethical sphere the counterpart of this 
nihilism would be the doctrine that there is no natural 
distinction between right and wrong. Plato, however, is 
very careful not to represent Gorgias as drawing this con- 
clusion himself, and even his ardent disciple Polos shrinks 
from the extreme consequences of opposing natural to 
legal right. These are drawn by one Kallikles, who is 
introduced as an Athenian democratic statesman. We 
know nothing of him otherwise, but he impresses us as a 
real man of flesh and blood. He is still young in the 
dialogue, and he may very well have disappeared during 
the revolutionary period. It is not Plato's way to introduce 

1 The title cannot be ancient in this form, as is shown by the use of 
to introduce an alternative. 


fictitious characters, nor does he introduce living con- 
temporaries, except where, as in the Phaedo^ that is made 
necessary by historical considerations, In any case, we 
have abundant evidence that the doctrine upheld by 
Kallikles, namely, that Might is Right, was current at 
Athens towards the close of the fifth century. In the 
Melian dialogue, Thucydides has shown us how it might 
be used to justify the attitude of the imperial democracy 
to its subject allies, and the Herakles of Euripides is a 
study of the same problem. 1 Its theme is that the 
"strong man" is not sufficient for himself, and is only 
safe so long as he uses his strength in the service of man- 
kind. This conception of the "strong man" (of which 
Herakles was the regular type) was not in itself an ignoble 
one. It had its ideal side, and Pindar sings how Herakles 
took the oxen of Geryones without paying for them in 
virtue of that higher law, which "justifies even the most 
violent deed with a high hand," a passage duly quoted 
in Plato's Gorgias (484 b). Such theories are a natural 
reaction against that rooted jealousy of everything above 
the common which is apt to characterise democracy. In 
modern times Carlyle and Nietzsche represent the same 
point of view. The worship of the strong man or " hero/* 
who can rise superior to all petty moral conventions 
in fact, of the " superman " seems to have been fostered 
in the fifth century B.C. by much the same influences as in 
the nineteenth century A.D. It is clear, then, that even 
the doctrine of Kallikles is not a complete ethical nihilism. 
Might really is Right. That is a very different thing 
from saying Right is Might. 

In the Republic that is the doctrine maintained by 
Thrasymachos. According to him there is no Right at 
all, and what we call by that name is only " the interest of 
the stronger" which he is able to force the weaker to 
accept as lawful and binding on themselves in virtue of his 
strength. It is important to observe that Thrasymachos 

1 See my paper "The Religions and Moral Ideas of Euripides/' in the 
Proceedings of the Classical duociation of Scotland, 1907-8, pp. 96 sqq. 


belongs to the generation we are now considering ; for 
readers of the Republic are often led to suppose, by an 
illusion we shall have to note more than once, that 
Plato is there dealing with the controversies of his own 
day. It is well to remember, then, that Thrasy machos 
was mentioned as a celebrated teacher of Rhetoric in 
the earliest comedy of Aristophanes, which was produced 
in 427 B.C., the year Plato was born and Gorgias came to 
Athens. It is not to be supposed that he was still living 
when the Republic was written ; he belonged to a genera- 
tion that was past and gone. We can hardly imagine 
anyone maintaining such vigorous doctrine in Plato's day, 
but it was natural enough that it should find advocates in 
the second half of the fifth century. It is the real ethical 
counterpart to the cosmological nihilism of Gorgias. 

Plato's final judgment on the Sophists (in the sense in 
which we have been using the word) is to be found in the 
Laws (889 e). It is that, by thus insisting on the oppo- 
sition between Law and Nature, they tended to do away 
with the distinction between right and wrong. If that 
distinction is not rooted in nature, but depends solely 
on human laws and institutions, it is valid only so long as 
we choose to recognise it. On the other hand, if we 
appeal from human law to a supposed higher law, the law 
of Nature, all restraint is abolished. We are forbidden 
by Plato's own account of them to attribute immoral 
intentions of any kind to the great Sophists ; but we can 
hardly dispute his estimate of the inevitable consequences 
of their teaching in a state of society such as existed at 
Athens in the closing decades of the fifth century. It is 
an impartial historical judgment ; for, in Plato's day, there 
were no longer any Sophists in the proper sense of the 

Eclectics and Reactionaries. 

98. Besides these men there were a good many 
others, also called " Sophists " by their contemporaries, 
who attempted to carry on the traditions of the Ionian 


cosmological schools. They were not, certainly, men of 
the same distinction as Protagoras or Gorgias, but they 
have their place in history as the vehicles by which 
the ideas of Ionian science were conveyed to Sokrates and 
his circle. From this point of view the most important of 
them is Diogenes of Apollonia, whose date is roughly 
fixed for us by the statement of Theophrastos that he 
borrowed from Anaxagoras and Leukippos, which shows 
that he belonged to the latter part of the fifth century 

We have considerable fragments of Diogenes, written 
in an Ionic prose similar to that of some of the Hippo- 
kratean writings. We find here the first explicit justifica- 
tion of the old Milesian doctrine that the primary substance 
must be one, an assumption which the rise of pluralism 
had made it necessary to defend. The action and reaction 
of things on one another, he says, can only be explained 
in this way. We may also trace the influence of Anaxa- 
goras in another matter. Diogenes not only said the 
primary substance was a " god," which was nothing new, 
but also identified it with Mind (vovi). On the other 
hand, he follows Anaximenes in holding that this primary 
substance is air, and in deriving all things from it by 
rarefaction and condensation. It is possible to see the 
influence of Herakleitos in the close connexion he 
established between wisdom and the dryness of the air we 
breathe. "Damp hinders thought" was one of his dicta, 
and is burlesqued in the Clouds (232) accordingly. In 
one respect only does Diogenes appear to have shown 
some originality, and that was in his medical work. His 
account of the veins was celebrated, and bears witness 
to the influence of Empedokles. 

Hippon of Samos is of less importance. He revived 
the doctrine of Thalcs that water was the primary sub- 
stance, and defended it on physiological grounds. We 
now know from Menon's latr'tka that he was a medical 
writer and that he was a native of Kroton. He was, 
therefore, one of the men who brought Western medicine 


to Ionia, and that accounts for the character of the argu- 
ments with which he defended his thesis. It is probable 
that the reasoning conjecturally attributed to Thales by 
Aristotle is really his. We may be sure that Thales 
defended his theory on meteorological, not physiological, 
grounds. That is just the difference between the two 

Archelaos of Athens was a disciple of Anaxagoras, and 
the first Athenian to interest himself in science or philo- 
sophy. He deserves mention for this, since, with the 
exception of Sokrates and Plato a considerable exception 
certainly there are hardly any other Athenian philosophers. 
There is not the slightest reason to doubt the statement 
that he had Sokrates for a disciple. The contemporary 
tragic poet, Ion of Chios, said in his Memoirs that 
Sokrates came to Samos in the company of Archelaos as 
a young man. We know that Ion gave an account of 
the visit of Sophokles and Perikles on the occasion of the 
blockade of Samos in 441/0, and this statement will refer 
to the same occasion. 1 Sokrates would be about twenty- 
eight at the time. Aristoxenos, as usual, repeats scandals 
about Archelaos and Sokrates. We are not bound to 
believe them, but they would have been pointless unless 
Sokrates had been generally known to have associated 
with Archelaos. Aristoxenos says that he was seventeen 
years old when this association began, and that it lasted 
many years. 2 Though Plato does not mention Archelaos 
by name, he refers unmistakably to his doctrines as having 
occupied Sokrates in his early youth, and it is natural 
to suppose that the man who is mentioned as reading 
aloud the book of Anaxagoras was no other than his 

1 Ion, fr. 73 (Kfipke). The title of Ion's work was 
("Visits"). There is no inconsistency between his statement and that 
of Plato (Crito, 52 b) that Sokrates never left Athens except on military 
service. This is a case of military service like the others we shall have 
to consider directly. It is most unlikely that Ion should have meant any 
other Sokrates in this connexion, as has been suggested. 

2 Aristoxenos, fr. 25 (F.H.G. ii. 280). 


Athenian disciple. 1 It is, therefore, quite unjustifiable 
to discredit the statement that Sokrates was his follower. 
It rests on practically contemporary evidence, and 
Theophrastos accepted it. 2 

1 Pkaedo, 96 b, 97 b, with 1117 notes. The theory that the warm and 
the cold gave rise by " putrefaction " (enjare&Sv) to a milky slime (tAvs), 
by which the first animals were nourished, is that of Archclaos, and is 
mentioned first among the doctrines Sokrates considered. 

Op. fr. 4 (Diels). 



The Problem 

99. It is possible to construct a biography of Sokrates 
from the dialogues of Plato, and, on the face of it, they 
seem to present us with an intelligible and consistent 
account of the man and his ways. Xenophon has left us 
three or four works purporting to record actual conversa- 
tions of Sokrates, whom he had known as a young man, 
but whom he saw for the last time just before he joined 
the expedition of Cyrus as a volunteer (401 B.C.). He 
tells us himself how he consulted Sokrates on the wisdom 
of that step, and was referred by him to the Delphic 
oracle. He was careful, however, not to ask the oracle 
whether he should join the expedition at all ; he only 
inquired to which of the gods he should offer prayer and 
sacrifice so as to ensure a prosperous issue to the journey 
he had in mind. He tells us frankly that Sokrates 
rebuked him for this evasion, and that is really all we 
know about their intercourse. If there had been much 
more to tell, we may be pretty sure Xenophon would 
have told it ; for he is by no means averse to talking 
about himself. At this time he was under thirty, and 
Sokrates had passed away before his return from Asia. 
Several of the Sokratic conversations he records are on 
subjects we know Xenophon was specially interested 
in, and the views put forward in them are just those 
he elsewhere expresses in his own name or through the 


mouth of Cyrus, the hero of his paedagogic romance. 
No one ever thinks, accordingly, of appealing to such 
works as The Complete Householder (the Oicoj/o/zi/co?) for 
evidence regarding " the historical Sokrates." There are 
two other writings, the Apology and the Symposium, which 
seem to have been suggested by the dialogues of Plato 
bearing the same names, and these are generally left out 
of account too. Since the eighteenth century, however, 
it has been customary to make an exception in favour of 
a single work, the Memorabilia, composed by the exiled 
Xenophon with the professed intention of showing that 
Sokrates was not irreligious, and that, so far from cor- 
rupting the young, he did them a great deal of good 
by his conversations. It is quite intelligible that the 
eighteenth century should have preferred the Sokrates of 
the Memorabilia to that of the Platonic dialogues ; for 
he comes much nearer the idea then current of what a 
philosopher ought to be. 1 In other respects it is hard to 
see what there is to recommend him. It is recognised 
that Xenophon is far from being a trustworthy historian, 
and the Cyropaedia shows he had a turn for philosophical 
romance. It is certainly unsound methodically to isolate 
the Memorabilia from Xenophon's other Sokratic writings, 
unless very strong reasons indeed can be given for doing 
so. Above all, it is quite impossible to get anything like 
a complete picture of Sokrates from the Memorabilia 
alone, and so in practice every writer fills in the out- 
line with as much of the Platonic Sokrates as happens 
to suit his preconceived ideas of the man. 2 Such a 

1 The first writer to prefer the Sokrates of the Memorabilia to the 
Platonic Sokrates was apparently Brucker (1741). The only reason he 
gives is that Xenophon had only one master, from whom he inherited 
not only moral philosophy, but integrity of life, while Plato was taken 
up with a " syncretism " of various doctrines. He quotes also an anecdote 
about Sokrates hearing the Lysis read, and observing, " Good heavens ! 
what lies the young man tells about me ! " But Sokrates was dead before 
the Lysis was written. 

2 In particular the " irony " of Sokrates comes entirely from Plato. 
The Sokrates of the Memorabilia has nq doubts or difficulties of any kind. 


procedure is hopelessly arbitrary, and can only land us in 
unverifiable speculations. It would be far better to say at 
once that we cannot know anything about Sokrates, and 
that for us he must remain a mere x. Even so, however, 
the Platonic Sokrates is actual enough, and he is the only 
Sokrates we can hope to know well. If he is a fictitious 
character, he is nevertheless more important than most 
men of flesh and blood. The only sound method, there- 
fore, is to describe his life and opinions without, in the 
first instance, using any other source. Only when we 
have done that can we profitably go on to consider how 
far the Sokrates we learn to know in this way will account 
for the slighter sketch of Xenophon. We shall also have 
to consider in what relation he stands to the caricature 
in the Clouds of Aristophanes. 

The Platonic Sokrates. 

100. Sokrates, son of Sophroniskos, of the deme 
Alopeke, was seventy years old, or a little more, when he 
was put to death (399 B.C.). 1 He was born, then, about 
470 B.C., some ten years after Salamis, and his early man- 
y hood was spent in the full glory of the Periklean age. His 
"family traced its descent to Daidalos, which means appar- 
ently that it was of some antiquity, and Sophroniskos 
must have been able to leave some property ; for we shall 
find Sokrates serving as a hoplite. His mother was a 
midwife, Phainarete by name, and she had another son, 
Patrokles, by another husband. It is worthy of note that 
the great Aristeides was of the same deme, and his son 
Lysimachos speaks of Sophroniskos in the Laches as a 
family friend. He says he never had any difference with 

1 Apol 1 7 d ; Crito, 520. We know the date of his death from Deme- 
trios Phalereus and the Marmor Parium. I have not given detailed 
references to the passages of Plato on which this account is based. They 
are well known and easily found. I do not think I have said anything 
which is not stated in Plato or to be immediately inferred from 
what Plato says. If this account of Sokrates is a " construction," it 
is Plato's, not mine. 


him to the day of his death. It is evident, then, that 
Sophroniskos was a man of some position In his deme. 
Another fellow-demesman was the wealthy Kriton, who 
was just the same age as Sokrates, and remained deeply 
attached to him till the end. 

Late in life Sokrates married Xanthippe, by whom 
he had three sons. When his father was put to death, 
the eldest of them, Lamprokles, was a lad ; but the other 
two, Sophroniskos and Menexenos, were children. The 
last named, indeed, was only a baby in arms. There 
,is no hint in Plato that Xanthippe was a shrew. Her 
name and those of her eldest and youngest sons suggest 
that she was a woman of good family. 1 In the Phaedo we 
are told that the friends of Sokrates found Xanthippe and 
her baby in the prison when the doors were opened. 
They must have passed the night there, and she was in an 
overwrought condition. Sokrates sent her home, but she 
returned later in the day with the other women of the 
family and spent some time with Sokrates in an inner 
room, where she received his final instructions in presence 
of the faithful Kriton. 2 

Sokrates was very far from handsome. He had a snub 
nose and strangely protruding eyes. His gait was peculiar, 
and Aristophanes likened it to the strut of some sort of 
waterfowl. In other places, his appearance is compared 
to that of a torpedo-fish, a Silenos, or a Satyr. He always 
went barefoot, save on special occasions, and he never 
went outside the town except on military service, and 
once to the Isthmian games. 

He was odd too in other ways. It was well known 
that, even as a boy, he had a "voice," which he called 
his " divine sign," and which he regarded as something 

1 It is noteworthy that It Is the second son who is called after the 
father of Sokrates. 

2 The scandal-monger Aristoxenos tried to fix a charge of bigamy 
on Sokrates. He said he was married at the same time to Xanthippe 
and to Myrto, the daughter of Aristeides. Ariste-Ides died in 468 B.C., 
so Myrto must have been about as old as Sokrates or older. 



peculiar to himself, and probably unique. It came to him 
often, and sometimes on the most trivial occasions. The 
remarkable thing about it was that it never prompted him 
to do anything ; it only opposed his doing something he 
was about to do. 1 Besides this, Sokrates was subject 
to ecstatic trances. He would stand still for hours together 
buried in thought, and quite forgetful of the outer world. 
His friends were accustomed to this and knew better than 
to disturb him when it happened. They simply left him 
alone till he came to himself. There was a celebrated 
occasion in the camp at Poteidaia, when Sokrates was 
not quite forty years old, on which he stood motionless 
from early morning on one day till sunrise on the next, 
buried in thought ((ppovrifyv r*) 5 as we are told in the 
Symposium. His comrades in arms were much astonished, 
and some of them brought their camp-beds into the open 
to see if he would really remain standing there all night. 
When the sun rose next morning, he said a prayer and 
went about his business. 2 

101. A man of this temperament would naturally 
be influenced by the religious movement of his time, and 
Plato indicates clearly that he was. He was a firm 
believer in the immortality of the soul and in the life 
to come, doctrines which were strange and unfamiliar 
to the Athenians of his day. He even believed, though 
not without reservations, in Rebirth and Reminiscence. 
When asked his authority for these beliefs, he would 
refer, not only to inspired poets like Pindar, but to " priests 
and priestesses who have been at pains to understand 
L the acts they perform."?, In particular he professed to 
have been instructed by a wise woman of Mantineia 

1 Xenophon makes a point of contradicting Plato as to this. He says 
the " voice " gave both negative and positive warnings. Obviously, if a 
young man asked Sokrates whether to go on a military adventure or not, 
and the " voice J> gave no sign, that could be interpreted as positive advice 
to go. The pseudo-Platonic Theages throws much light on the subject. 

2 Symf. 220 c-d. The statement would be pointless if it were not true. 

8 1 a. 


named Diotima. To the very end of his life, he was 
deeply interested in what he called "sayings of yore" 
or the "ancient word," and expressly attributed to 
Orpheus, 1 according to which the body is a tomb in 
which the soul is kept in custody. It cannot attain to 
perfect purity till it is released from the body by God, 
whose chattel it is, and comes to be alone by itself. Then, 
and not till then, can it dwell with God. The man who 
L follows philosophy, which is the highest music, will there- 
fore practise death even in his lifetime by accustoming his 
soul to concentrate upon itself, and so to attain such wisdom 
as may be possible in this world. 

But, with all this, Sokrates was no mere visionary. He 
had a strong vein of shrewd common sense that kept him 
u from committing himself to the often fantastic details of 
Orphic and Pythagorean religion, however powerfully 
these might appeal to his imagination. He calls the 
doctrine that the soul is imprisoned in the body, a "high 
one and not easy to understand," and though he was 
certain that the souls of the righteous ^would be with God 
when they departed from the body, he could not feel equally 
.sure that they would be, with the saints. When he related 
eschatological myths in the Orphic style, as he often did, 
Jie used to warn his hearers that they were at best some- 
thing like the truth. No man of sense would insist on 
their literal accuracy. Besides this, he had a healthy 
contempt for the common run of Orphic and other 
traffickers in pardons and indulgences, whom he accused 
of demoralising the nation by their gross descriptions of 
heavenly joys. That, however, was perfectly consistent 
with the belief that Orphicism contained, in however dim 
u a form, a great truth not to be found in the ordinary 
religion of the State. The manner of its expression he 
compared to fables or riddles, of which not everyone 
can guess the true sense. 

1 02. The truth is that there were two well-marked 
sides to his character. He was indeed a visionary or 

1 Crat. 400 c. 


4< enthusiast/' in the Greek sense of that word, but he was 
also uncommonly shrewd. His critics called him " sly/' 
using a word (e?/>o)v), which is properly applied to foxes. 
The Scots word " canny" (not always a term of praise) 

, comes nearest in meaning to the Greek. He did not like 
to commit himself further than he could see clearly, and he 
was apt to depreciate both his own powers and other 
people's. That was not a mere pose; it was due to an 
instinctive shrinking from everything exaggerated and 

^ insincere. As has been indicated, it is only the opponents 
of Sokrates that charge him with "irony" (efyxtWa), a 
word which undoubtedly suggested the idea of humbug; 
but Plato shows us over and over again the real trait in 
his character which this uncomplimentary description was 
aimed at, with the result that the word "irony" has 
changed its meaning for us. To a very large extent, we 
gather, " the accustomed irony" of Sokrates was nothing 
more or less than what we call a sense of humour which 
enabled him to see things in their proper proportions. 

103. His interest in religion of a mystic type would 
naturally lead Sokrates to seek light from the science of his 
time. The two things were very closely connected at this 
date, as we have seen when dealing with Empedokles. In 
the Phaedo (96 a sqq.} Plato makes Sokrates give an account 
of his intellectual development which must be intended 
to be historical, seeing that the questions described as 
occupying his mind are just those that were of interest 
at Athens when Sokrates was a young man, and at no 
other time or place. 1 He asked himself whether life had 

1 For a detailed discussion of these see the notes in my edition of the 
PhacdO) ad loc. The main point is that Sokrates is represented as hesitat- 
ing between Ionic doctrine, such as he would learn from Archelaos and 
Diogenes (cp. 93), and Italic doctrines, some of which belong to the 
school of Empedokles, whilst others are Pythagorean. Sokrates may 
have learnt the latter directly or indirectly from Philolaos. Empedokles, 
who took part in the colonisation of Thourioi, probably visited Athens 
(for we know that Kritias adopted his theory of sensation) and it is not 
difficult to suppose that Philolaos came there too. Athens is the only 
place where the Ionic and Italic philosophies could come into sharp 


arisen from the putrefaction of the warm and the cold 
(a doctrine we know to have been that of Archelaos), 
and whether the earth is flat (as the lonians taught) or 
round (as the Pythagoreans held). He was interested in 
the relation between sensation, belief, and knowledge (a 
problem raised by Alkmaion), and he considered whether 
"what we think with" is air (the doctrine of Diogenes) or 
blood (that of Empedokles). In fact, he is represented as 
having been influenced by practically every theory repre- 
sented at Athens in the middle of the fifth century. But 
none of these could give him satisfaction ; for they threw 
no light on what he chiefly wanted to know, the cause of 
things, why things are what they are and become what 
they become. They explained everything mechanically, 
whereas Sokrates wished to be shown that everything is 
as it is because it is best for it to be so. The system of 
Anaxagoras, indeed, seemed more promising at first; for 
it attributed the origin of the world to Mind. But this 
proved disappointing too ; for Anaxagoras made no use of 
Mind except when he was at a loss for another explanation. 
Otherwise he spoke of " airs" and " aethers" just like the 
rest. Sokrates accordingly turned his back on all such 
speculations, and resolved to work out a new method for 

104. According to Plato, Sokrates must have reached 
this point when he was quite young; for he makes him 
discuss his new theory with Parmenides and Zeno when 
they visited Athens shortly after the middle of the century 
( 63). It is also made clear that he came into contact 
with the great " Sophists" of the day at a very early age. 
The first visit of Protagoras to Athens must have taken 
place before Perikles entrusted him with the important 
duty of legislating for Thourioi in 444 B.C., that is to say, 
it must have coincided very nearly with the visit of Par- 
menides and Zeno, and we have seen that tradition repre- 
sents Zeno and Protagoras as engaged in controversy. On 

conflict like this, and the middle of the fifth century is the only time at 
which it could happen. 


his second visit, several years later, Protagoras remembers 
the young Sokrates quite well. He is made to say that of 
all the people he meets he admires Sokrates most, certainly 
far more than anyone else of his age. 1 A very similar 
compliment is put into the mouth of Parmenides. 2 Plato 
clearly means us to understand that Sokrates had attracted 
the notice of the most distinguished men of the time when 
he was not more than about twenty-five. 8 He was also 
intimate with Hippias and Prodikos, and he used to say 
that he had attended one of the cheaper courses on 
synonyms given by the latter. Gorgias, on the other 
hand, did not visit Athens till Sokrates was over forty 
years old. 

It is clear, however, that Zeno, " the Eleatic Palamedes,"* 
had more influence on Sokrates than anyone. As Aristotle 
said, 6 he was the real inventor of Dialectic, that is to say, 
the art of argument by question and answer. If the Peri- 
klean age had left any literature we should probably hear 
more about his work at Athens than we do, but the 
Athenians of the middle of the fifth century did not write 
books. We have traces enough, however, of the impression 
he left. We are told in the Parmenides of young Athenians 
who had been his associates, and it is recorded that Perikles 
himself " heard " him ( 63). We shall see that the Eleatic 
philosophy was sedulously cultivated at Megara, where its 
dialectical side was still further developed. Dialectic is 
literally the art of conversation or discussion, and its pro- 
cedure is governed by strict rules. The " answerer " 
(o aTTOKpivojuLGvoi) is required to reply to the questioner (6 

l Prot. 3610. Protagoras adds that he would not be surprised if 
Sokrates became distinguished for wisdom. Surely that is the remark of 
an old man to a very young one, not that of a man under sixty to a man 
over forty. Cp. 89. 

2 Parm. 1303. Cf. tb. I3$d. 

8 This is strikingly confirmed by the statement of Aristoxenos tnat 
Sokrates became a disciple of Archclaos at the age of seventeen (p. 124, . 2). 

iPhaedr 261 d. 

* In his dialogue entitled the Sofhist (aj>. Diog. Laert. ix. 20. 


in the fewest possible words, and to answer the 
question exactly as it is put. He is not allowed to ask 
other questions or to boggle at the form of , those put to 
him. Obviously this is a procedure which can be employed 
in the most fallacious manner, and in the Euthydemus we 
have a delightful sketch of its abuse. Even that, however, 
was of service in directing attention to the nature of the 
most common fallacies, and this helped in turn to indicate 
the direction in which the real difficulties were to be looked 
for. At any rate, it was the method that appealed most to 
Sokrates, and there can be little doubt he learnt it from 
Zeno. The influence of Zeno is also attested by the 
Phaedo (96e), where Sokrates is represented as puzzled, 
not only by the problem of growth, which was that of 
Anaxagoras and Archelaos, but also, and even more, by 
that of the unit, which was the special object of Zeno's 

105. If we bear in mind the extreme youth of 
Sokrates when he began to strike out a line for himself, 
and also how unusual it was for an Athenian to busy him- 
self seriously with such matters, we shall not be surprised 
to find that he had enthusiastic admirers among the 
younger men. We see from the opening scene of the 
Protagoras how some of them looked up to him as a guide 
even then, and consulted him about their studies. One 
of these, Chairephon, was particularly enthusiastic, and 
actually asked the Delphic oracle whether there was 
anyone wiser than Sokrates. The Pythia of course replied 
that there was no one. That proved a turning-point in 
the life of Sokrates, but Plato is careful to let us know that 
he did not accept the oracular response at its face value. 
His humour (elpoov&a) did not fail him when he turned it 
on himself, and he at once set out to prove the god in the 
wrong. He would find someone wiser than himself, and 
use him to refute the oracle. So he went to one of the 
politicians, whose name he does not think it necessary to 
mention, and talked to him, with the result that he found 
him wise, indeed, in his own opinion and that of other 


people, but really quite ignorant. And he had the same 
experience w'th one set of people after another. The 
poets could give no intelligible account of their own 
works. Apparently it was by some sort of divine inspira- 
tion they succeeded ; for they did not know how it was 
themselves. The craftsmen, indeed, did as a rule know 
something about their own trades, but unfortunately, on 
the strength of this bit of knowledge, they fancied they 
knew a great many other things of which they were quite 
ignorant, such, for instance, as how to govern an empire. 
At last he saw what the god meant. Neither Sokrates 
nor anyone else knew anything, but Sokrates was wiser 
than other men in one respect, namely, that he knew he 
was ignorant and other men did not know they were. 
From this time forward, he regarded himself as having a 
mission to his fellow-citizens. He had been set apart 
by God to convince them of their ignorance. 

Now according to Plato all this happened before the 
beginning of the Peloponnesian War ; for Sokrates is 
represented as resuming his mission after his return from 
Poteidaia. 1 We cannot, therefore, date the oracle later 
than about his thirty-fifth year, and it is obvious that he 
was already well known by that time. The inquiry of 
Chairephon would be inexplicable on any other supposi- 
tion. Plato himself was not born yet, and of course what 
he tells us must be based on the statements of Sokrates 
himself, and no doubt of Chairephon. It does not require 
great literary tact to see that Sokrates only took the oracle 
half-seriously, and that what he did was to apply to it the 
same methods of interpretation that he usually applied to 
Orphic and other mythology. On the other hand, he 
clearly believed it quite possible that a higher power 
might make use of oracles, dreams, and the like to com- 
municate with human beings. He was the least dogmatic 
of men on such subjects, and his own " voice" and his 
visions seemed a case in point. What is quite certain is 
that he sincerely believed his mission to be imposed on 

1 Charm, 153 a. 


him by God. He gave up everything for it, and that 
was the cause of his poverty in later life. He spoke of his 
service (\arpeia) to God 3 and called himself the fellow-slave 
(ojuo&wAoff) of Apollo's swans. That, according to Plato, 
was a genuine faith, and he was intensely in earnest about it. 

1 06. The mission of Sokrates was interrupted by the 
outbreak of the Peloponnesian War, in which he was 
called on to do his duty as a citizen-soldier. He fought 
at Poteidaia (432 B.C.), at Delion (424 B.C.), and at 
Amphipolis (422 B.C.), and Plato has been careful to leave 
a record of his bravery in the field. 1 In the Symposium 
(220 d sg.*) he makes Alkibiades describe his conduct 
with enthusiasm. In one of the battles Alkibiades was 
wounded, and Sokrates saved his life by watching over 
him till the danger was past. The generals awarded the 
prize of valour to Alkibiades, but he himself maintained 
it ought to go to Sokrates. Again at Delion, when the 
Athenians had to retreat, Alkibiades tells how Sokrates 
retired along with Laches, and far surpassed him in 
presence of mind, so that they both came off unhurt 
Laches is made to refer to the same incident in the 
dialogue called by his name (181 b), and he adds that, 
if everyone else had done his duty like Sokrates, the 
defeat would have been turned into a victory, Sokrates 
was then about forty-six. 2 

107. As we shall see, he had by this time gathered 
round him a circle of associates (eraipoi), but these must 
be carefully distinguished from the young men he 
influenced in the course of his public mission. It appears, 
in the first place, that he exercised a singular fascination 
over those who were devoting themselves to what was 

1 We have seen ( 98) that lie probably served at Samos in 441/0,, 
but Plato has no occasion to mention that. It was before the time of 
most of the speakers in his dialogues. It is interesting to think that 
Sokrates fought against a force commanded by Melissos. 

2 It is important to notice the way Plato insists on the military 
reputation of Sokrates, It accounts for the interest taken in him by 
Meno, Xenophon and others at a later date. See my edition of die 
Phaedo (Introduction, p, xiv). 


then the new calling of a professional soldier. That was 
only natural, and in the Republic Plato represents Sokrates 
as strongly impressed by the necessity for a professional 
army. Besides these there were, we are told, a number 
of young men of good family, who had no profession on 
which they could be cross-examined, and who took great 
pleasure in hearing the ignorance of others exposed. Some 
of them even thought they might get a better preparation 
for public life by listening to Sokrates than any professional 
Sophist could give them. It is certain that Kritias asso- 
ciated with Sokrates in this way, though he did not do so 
for long. We hear of others, such as the fellow-demes- 
man of Sokrates, Aristeides, son of Lysimachos, who soon 
fell away. No doubt they wished to learn the art of suc- 
cess, whereas Sokrates insisted on the necessity of serious 
study for a politician, just as for any other craftsman. 
There were others who were really devoted to him, notably 
Alkibiades and Charmides. Charmides was Plato's uncle, 
and it was doubtless through him that Plato came to 
associate with Sokrates. Even these, however, are not to 
be regarded as his disciples, or even as his associates in the 
strict sense like Chairephon. In the Apology he speaks of 
them as " those they say are my disciples." x 

108. In speaking of his relations with these young 
men Sokrates habitually used the language of love, 
tempered, of course, by his usual sly humour. To under- 
stand this, we must remember that at Thebes and Elis 
and in the Dorian States attachments of this kind were a 
recognised institution. They had their origin in the 
romantic relation of knight, squire and page in the Greek 
Middle Ages, and they were believed to have great value 
for military purposes. 2 In the Laws (636 b jy.) the 

^ Apol. 33 a. In his Bousirls (u. 5) Isokrates represents the matter 
exactly as Plato makes Sokrates represent it himself. He criticises Poly- 
krates (Cf. 116, infra) for making Alkibiades a disciple (/x-a^rrfc) of 
Sokrates, whereas no one ever knew of him being educated (T 
by Sokrates. 

2 See Bethe in Rhein. Mus. Ixii. (1907), pp. 438 sqq. 


Athenian Stranger, that is to say Plato, criticises the 
institutions of Sparta and Crete on the very ground that 
they were favourable to the abuse of such relationships. 1 
In the Ionian States generally, on the other hand, they 
were considered disgraceful, 2 and, though the Dorian 
custom had made its way into Athens before the time 
of Solon, its abuse was condemned both by law and by 
public opinion. 3 Plato makes it abundantly clear, how- 
ever, that it was the fashion in aristocratic circles to ape 
this feature of Spartan life among others. If we may 
trust the extremely vivid account of the matter he puts 
into the mouth of Alkibiades and it is surely incredible 
that he invented it it was Alkibiades himself that first 
posed as the epwj&evo? of Sokrates, though it is also made 
quite clear that it was only a pose. The personal chastity 
of Sokrates is assumed as the foundation of the whole 
story, and we have therefore no right to interpret his 
language in a gross sense. What really surprises a modern 
reader is the matter-of-fact way in which the abuse of such 
relationships is spoken of. It will help us to understand 
that, if we remember that at Megara, only a few miles 
from Athens, no disgrace attached to it. In these circum- 
stances, we can hardly look for the same reticence on the 
subject as is commonly observed at the present day, though 
Plato's condemnation is unequivocal. 

The thing appealed to Sokrates on another side, how- 
ever, and here we may note once more his accustomed 
humour. He had a way of speaking of the birth of 
thoughts in the soul in language derived from his mother's 
calling. He professed, of course, that he himself was 
incapable of giving birth to wisdom, but he claimed to 
be an excellent man-midwife, well skilled in the art of 

1 Addressing a Spartan and a Cretan, he says: /ecu rourov ras 

S TToAetS TT/XOTttS CtV TIS atTt<J)TO (636 b). 

2 Plato, Symf. 182 b. 

8 Plato, Phaedr. 231 e : et TOLVVV rbv vo/iov rov Aca^ecrr^fcora 
fj.rj TTvOojjLtvw TO>V dv$/oa>7ra>i' eifetSos crot, yevrjrai KT\, Aischines 
Against Timarchos, passim. 


bringing new thoughts to the birth. Besides that, just as 
midwives are the best matchmakers, he claimed to have 
a peculiar gift for discerning who the best teacher for a 
young man would be. That is all playful, to be sure, 
but we must never forget that Sokrates was a mystic as 
well as a humorist, and the mystics have always found 
the language of love more adequate than any other to 
express their peculiar experience. The love of a fair body 
is only the earthly type of something far higher. It leads 
on to the love of a fair soul, to the love of fair studies and 
fair ways of life, and at last it brings us into the very 
presence of the "forms" of beauty, righteousness, and 
holiness in that supercelcstial region where they have their 
dwelling-place. 1 When thus regarded as the objects of 
love, these " forms " are seen to be the realities of which 
the things in this world arc but shadows, and from which 
they derive such imperfect being as they have. There can 
be no doubt Plato means us to believe that Sokrates 
had actually attained to this beatific vision. It is not for 
nothing that he is represented as having one of his trances 
just before the conversation recorded in the Symposium. 
That must be intended to throw light on that other trance 
of twenty-four hours in the camp at Poteidaia more than 
a dozen years before. The man who saved the life of 
Alkibiades by his fearless devotion in the battle was fresh 
from the contemplation of a far higher beauty than his. 

109. Plato has left us more than one description of 
the effect the discourses of Sokrates had on young men. 
It will be well to quote the words he puts into the mouth 
of Meno, a reluctant admirer, and Alkibiades, an enthu- 
siastic one. Meno says (Meno> 79 e) : 

Before I met you I was told you did nothing but confuse your- 
self and make other people confused. And now I really think 
you arc just bewitching me and casting spells and enchant- 

1 Phaedr. 247 c sqq. I cannot believe that this is a description of 
Plato's own experience. It is strictly in keeping with all we know 
about the temperament of Sokrates. 


ments over me, so that I am full of confusion. I think, if 
I may be allowed the jest, you have a strong resemblance, 
not only in figure but in other respects, to the torpedo-fish. 
It benumbs anyone who comes near it and touches it, and 
that is just what you have done to me. Both my soul and 
my lips are literally benumbed, and I don't know what answer 
to give you. I have made speeches over and over again about 
goodness, and before large companies, with complete success 
as I fancied, but now I can't even tell what it is. I think it 
extremely prudent on your part never to take a voyage or 
leave your own country. If you were to do these things as a 
stranger in a foreign land, you would probably be taken up 
for a sorcerer. 

And Alkibiades, who, with all his faults, or because of 
them, was very dear to Sokrates, says this (Symp. 215 a) : 

I shall endeavour to praise Sokrates as well as I can by 
means of images. Very likely he will think it is to make 
fun of him, but my image is chosen for its truth and not its 
absurdity. I say he is just like the figures of Silenos we see 
in the statuaries' shops, those they make with pipes or flutes 
in their hands, and when you open them you find they have 
images of the gods inside them. And I say too that he is like 
the satyr Marsyas. That you are like these in appearance, 
Sokrates, I fancy you won't deny yourself, and now let me 
tell you how you are like them in other ways. You're a 
wanton, aren't you ? If you don't admit it, I shall call wit- 
nesses. Ay, and aren't you a piper ? A far more wonderful 
one than he was ! He only charmed men by his instruments ; 
* , . you beat him because you produce the very same effect 
by words alone without any instrument. When we hear any- 
one else speak, even a very good speaker, none of us care a 
bit ; but when anyone hears you or anyone else repeating 
your words, even if the speaker is an indifferent one, and 
whether it is a woman or a man or a lad that hears him, we 
are all confounded and inspired. My friends, unless I was 
afraid you would think me quite drunk, I would tell jou on 
my oath the effect his words have had on me and still have. 
When I listen to him my heart leaps even more wildly than 
those of people in a Korybantic ecstasy, and his words make the 
tears gush from my eyes. And I see many others affected in 
the same way. When I used to hear Perikles and other good 
speakers, I thought they spoke very well, but I had none of 


these feelings. My soul was not troubled or angry at the 
idea that it was in a state like a slave's. But I have often 
been put into such a condition by this Marsyas here, that 
I thought life not worth living so long as I remained as I was. 
And I am quite sure that if I were to consent to lend him my 
ears now, I couldn't hold out, but should feel just the same. 
He forces me to confess that, though I myself fall far short in 
many a thing, I neglect myself and busy myself about the 
affairs of Athens. So I stop my ears and run away from him 
as if from the Sirens, to prevent myself becoming rooted to 
the spot and growing old by his side. Why, he is the only 
human being that has ever made me feel ashamed in his 
presence, a feeling of which I might be supposed incapable. 
I know very well I can give no reason for not doing what he 
tells me to, but, when I have left him, I find my popularity 
too much for me. So I act like a runaway slave and a fugitive, 
and whenever I see him, I am ashamed of the admissions I 
have made. Many a time I feel that I should be glad to see 
him wiped out of existence altogether, and yet, if that were 
to happen, I know I should be far more distressed than relieved. 
In fact I don't know what to make of him. 

Of course Plato himself was too young to hear Alkibiades 
talk like that, but he had opportunities enough of knowing 
about his relations to Sokrates. It is at least plain that he 
believed Sokrates to have been capable of exerting this 
fascination over Alkibiades as late as 416 B.C., when the 
banquet described in the Symposium is supposed to take 
place. It is natural, too, to regard the passage as evidence 
of the effect produced by the discourses of Sokrates on 
Plato himself in his youth. 1 

> no. In 423 B.C. Aristophanes produced the Clouds^ in 
which Sokrates, then about forty-seven years old, was the 
central figure. It will be necessary to say something later 
as to the picture there drawn of him ; here we have only 
to do with what Plato says about it. It is true that, in the 
Apology^ he makes Sokrates attribute much of the popular 
prejudice against him to the Clouds. He had been repre- 
sented as walking on air and talking a lot of nonsense 

1 It is not easy to imagine such discourses as we find in Xenophon's 
Memorabilia producing such effects as these 


about the things in the heavens and those beneath the 
earth, and that, he says, suggested the notion that he was 
irreligious. It may very well have done so at the time of 
his trial, when old memories of the Clouds would occur to 
the judges in confirmation of the charges Sokrates had 
then to face, but we gather also from Plato that no one 
took it very seriously at the time, least of all Sokrates and 
his circle, In the Symposium ', Sokrates and Aristophanes 
are represented as the best of friends six or seven years 
after the production of the Clouds^ and Alkibiades does not 
hesitate to quote a burlesque description of the gait of 
Sokrates from that very play. We are to understand, then, 
that at the time no offence was taken, and we need not 
suppose any was meant. It was only in the light of sub- 
sequent events that the Clouds was resented, and even so 
the matter is quite lightly treated in the Apology. 

111. But more difficult times were at hand. We have 
seen that Sokrates did his duty as a soldier, but he never 
held any office. The " voice " would not allow him to 
take part in politics. In 406 B.C., however, it fell to 
his lot to be a member of the Council of Five Hundred, 
and it so happened that it was the turn of the fifty 
representatives of the tribe Antiochis,, to which his deme 
belonged, to act as the executive committee of the Council 
at the time the generals were tried for failing to recover the 
bodies of the dead after the naval battle of the Arginoussai. 
The conduct of the trial showed that the democracy was 
getting into an ugly temper. It was proposed to judge 
all the generals together instead of taking the case or each 
separately. That was against the law, and Sokrates, who 
presided, refused, in spite of the .popular clamour, to put 
the question to the meeting. The generals were ultimately 
condemned by an illegal procedure, but the action of 
Sokrates made a deep impression, and he referred to it 
with justifiable pride at his trial. A little later, during 
the illegal rule of the Thirty, he had the opportunity of 
showing that he could not be intimidated by the other 
side either, The Thirty sent for him along with four 


others and gave them orders to arrest Leon of Salamis 
that he might be put to death. The four others carried 
out the order, but Sokrates simply went home. Plato 
makes him say that he would probably have suffered for 
this if the Thirty had not been overthrown shortly after. 
From this we may infer and we shall see that the point 
is of consequence that Sokrates did not feel called upon 
to leave Athens with the democrats, though his devoted 
disciple, Chairephon, did so. 

Aristophanes and Xenophon. 

112. Let us now consider how far this account of 
Sokrates is confirmed or otherwise by Aristophanes and 
Xenophon. In the first place, we must observe that Plato 
represents the life of Sokrates as sharply divided into two 
periods by the response of the oracle. In the earlier, 
he was chiefly occupied with the religious and scientific 
movements of his time, and with his new theory of the 
participation of sensible things in the " forms"; in the 
latter, his mission to his fellow-citizens is his chief, and 
almost his sole interest, though in the month that elapsed 
between his condemnation and his death he naturally 
recurred to the themes that had busied his youth. It is 
further to be noticed that the testimony of Aristophanes 
refers to the first of these periods, and that of Xenophon 
to the second. The Clouds was produced in 423 B.C., the 
year between the battles of Delion and of Amphipolis, 
in both of which Sokrates fought. His mission, though 
begun, was interrupted, and Aristophanes would be think- 
ing mainly of the earlier Sokrates. Chronology is vital in 
dealing with this question, and we must never allow our- 
selves to forget that Sokrates was only forty-seven when 
Aristophanes produced the Clouds, and that Plato and 
Xenophon were babies. We must, therefore, compare the 
caricature of Aristophanes only with what Plato tells us 
of the youth of Sokrates, and not with what he tells us 
of the later period. 


113. That the Clouds is a caricature is obvious, and 
it must be interpreted accordingly. There are two canons 
for the interpretation of comedy which are often neglected. 
In the first place, the very occurrence of a statement in a 
comedy affords a presumption that it is not a mere state- 
ment of fact. Statements of fact are not funny. On the 
other hand, every such statement must have some sort of 
foundation in fact ; for absolute fictions about real people 
are not funny either. What we have to ask, then, is 
what Sokrates must have been in the earlier period of his 
life to make the caricature of the Clouds possible. In the 
first place, he must have been a student of natural science, 
and he must have been interested at one time or other in 
the things in the heavens (ra fjLerewpa) and the things 
beneath the earth (ra VTTO yni). Plato makes Sokrates 
declare that these were the chief studies of his youth. 
Aristophanes represents Sokrates as an adherent of a 
system which is recognisable as that of Diogenes of 
Apollonia, and that is just why the chorus consists of 
clouds. We know that Diogenes had revived the theory 
of Anaximenes that everything is condensed or rarefied 
" air," and the clouds are one of the first results of the 
condensation of air. Just so Plato makes Sokrates say 
that he had studied, among other questions, whether 
" what we think with " was air (the doctrine of Diogenes) 
or blood (the doctrine of Empedokles), and Aristophanes 
represents him as swinging in a basket in order to get pure 
dry air for his thought. Aristophanes also knows of the 
spiritual midwifery of Sokrates, for he has a jest about 
the miscarriage of a thought. On the other hand, he 
represents him as a spiritualistic medium, and he calls 
the inmates of the Phrontisterion "souls," a word which 
to the ordinary Athenian would only suggest ghosts. 
He also ridicules them for going barefoot and unwashed, 
and speaks of them as "semi-corpses." All that, and 
more of the same kind, has a sufficient foundation in 
what Plato tells us of the Sokratic doctrine of the soul 
and the "practice of death." The only thing that strikes 


us at first as inconsistent with everything we can gather 
from Plato is that Sokrates teaches his pupils to make the 
weaker argument the stronger. That is not true even of 
Protagoras in the sense suggested, while the introduction 
of the Righteous and the Wicked Logos (possibly a later 
addition) seems even wider of the mark. And yet, if we 
look closer, we shall find there are sufficient indications 
of features in the teachings of the Platonic Sokrates to 
account for such a distortion on the part of a not too 
scrupulous comic poet. We know from Plato that the 
new method of Sokrates consisted precisely in the con- 
sideration of things from the point of view of proposi- 
tions (Ao'yo:) rather than from that of facts (ejoya), and 
Aristophanes would not be able, and certainly would not 
care, to distinguish that from the "art of Ao-you," which 
seemed so dangerous to conservative Athenians. As for 
the suggestion that it was used for the purpose of 
establishing immoral conclusions, we need only suppose 
that discussions like that described in the Hippias minor 
had got talked about, as they certainly would. It would 
seem obvious to the plain man that anyone who 
maintained the voluntary wrongdoer to be better than 
the involuntary must be engaged in the subversion of 
morality. I submit, then, that if the Sokrates of this 
date was much what Plato represents him to have been, 
the caricature of the Clouds is quite intelligible ; if he was 
not, it is surely pointless. 

114. But, above all, Aristophanes confirms Plato in 
the most explicit way by drawing a clear distinction 
between certain "disciples" (/xa^ra/), as he calls them, 
of Sokrates, of whom Chairephon was the chief, and 
who were his permanent associates (eraF/xn) in a scientific 
school, and the young men who frequented his society or 
were sent to him by their parents in order to learn how 
to succeed in life. What Plato tells us about Lysimachos 
and Aristeides 1 is enough to justify the burlesque figures 
of Strepsiades and Pheidippides. But the machinery of 

1 Laches, 1783 sqq. ; Theaet. 1 5 I a, 


the Phrontisterion implies that there was something much 
more serious. It is usually said, indeed, that Aristophanes 
is taking Sokrates as a type of the Sophists of the day, 
but that view is untenable. In the first place, the Old 
Comedy does not deal in types but personalities, and when 
Aristophanes does introduce a type, as in the Birds> he 
gives him a fictitious name. But apart from that, the 
Sophists of the day had no permanent associates. They 
were here to-day and gone to-morrow, and they only 
gave short courses of lectures to audiences that were per- 
petually changing. Besides, they were the last people in 
the world to trouble themselves with scientific inquiries 
such as Aristophanes is obviously making fun of. The 
Phrontisterion y in fact, is a burlesque of an organised 
scientific school of a type which was well known in Ionia 
and Italy, but had not hitherto existed at Athens, unless, 
indeed, Archelaos had established one. If Sokrates did 
not, in fact, preside over such a society, are we to suppose 
that Aristophanes himself invented the idea of a scientific 
school, or that he knew of those in other cities by hearsay 
and transferred them in imagination to Athens ? It is 
surely very hard to see what the point of that could be, 
and we must conclude, I think, that he expected his 
audience to know what an institution of the kind 
was like. If he has voluntarily or involuntarily con- 
fused Sokrates with anyone, it is not with Sophists like 
Protagoras and Gorgias or their followers, but with 
Anaxagoras and Archelaos ; and, if the latter did found a 
regular school, Sokrates would naturally succeed him as 
its head. That, in fact, seems to me the most probable 
account of the matter. We have seen that Sokrates was 
a disciple of Archelaos for a number of years. 1 

1 1 5. When we come to Xenophon, we must remember, 
in the first place, that he was very young, and Sokrates 
already an old man, when he knew him, and that he left 
Athens never to return about three years before Sokrates 
was put to death. In the second place, we must remember 

p. 124, n. 2, 


that the Memorabilia is an apologia , and must be judged 
by the canons of criticism applicable to such writings. 
The chief of these is that most weight is to be attached 
to statements not directly connected with the main pur- 
pose of the work ; above all, when they seem to involve 
admissions in any degree inconsistent with that. Now 
what Xenophon wished to prove is that Sokrates was 
unjustly accused of being irreligious, and that his conver- 
sations, so far from corrupting the young, did them a 
great deal of good. One of the chief arguments for the 
soundness of his religious attitude is that he refused to 
busy himself with natural science and dissuaded others 
from studying it. What Plato tells us of the disappoint- 
ment of Sokrates with Anaxagoras, and his renunciation of 
physical speculations at an early age, is enough to explain 
what Xenophon says, and yet he feels at once that he has 
gone too far. In fact he gives his point away completely 
by adding twice over : " Yet he himself was not unversed 
in these subjects" subjects of which he gives a list, and 
which correspond exactly to the most highly developed 
mathematics and astronomy of the time. 1 'Further, he 
knew that what Aristophanes burlesqued as the Phrontis- 
/m*0wwas a reality; for he makes Sokrates tell the Sophist 
Antiphon, who was trying to rob him of his disciples a 
very significant touch that he does in fact study the 
writings of the older philosophers with his friends. " I 
spend my time with them," he says, "unrolling the 
treasures of the men of old, which they have written 
down in books and left behind them." 2 Admissions like 
these are far more important than the philistine words put 
into the mouth of Sokrates about scientific study. No one 
who talked like that could have attracted Pythagoreans 
like Kebes and Simmias from Thebes to listen to him, as 
Xenophon also says he did. 8 

It would be possible to find a good many more admis- 
sions of this sort in Xenophon, but it is not clear to me 
how far the Memorabilia can be regarded as independent 

* Mem. iv. 7. 3-5. ^Mem, i. 6. 14. z Mem. iii. u. 17. 


testimony at all. In fact, it seems hardly possible 
to doubt that Xenophon got the greater part of his 
information about Sokrates from the dialogues of Plato. 
Otherwise, it would be very significant that he has heard 
of the importance of " hypothesis " in the dialectic of 
Sokrates. 1 I do not feel able to rely on such things 
as first-hand evidence, however, and therefore I make no 
use of them. Those who treat the Memorabilia as a 
historical work are bound, on the other hand, to admit 
a good many things that are hard to explain on the 
assumption that Sokrates was the sort of man Xenophon 
wishes us to think he was. In fact, Xenophon's defence 
of Sokrates is too successful. He would never have been 
put to death if he had been like that. 

1 1 6. The conclusion we are, in my opinion, forced to 
is that, while it is quite impossible to regard the Sokrates 
of Aristophanes and the Sokrates of Xenophon as the 
same person., there is no difficulty in regarding both as 
distorted images of the Sokrates we know from Plato. 
The first is legitimately distorted for comic effect ; the 
latter, not so legitimately, for apologetic reasons. To 
avoid misunderstanding, I should say that I do not regard 
the dialogues of Plato as records of actual conversations, 
though I think it probable that there are such embedded 
in them. I also admit fully that the Platonic Sokrates is 
Sokrates as Plato saw him, and that his image may to 
some extent be transfigured by the memory of his martyr- 
dom. The extent to which that has happened we cannot, 
of course, determine, but I do not believe it has seriously 
falsified the picture. Like Shakespeare, Plato had a 
marvellous gift of suppressing his own personality when 
engaged in dramatic composition. That is why his 
personality is so elusive, and why that of Sokrates has 
so often been substituted for it. We shall return to this 
when we come to Plato himself, but first I must warn the 
reader that there is another view of the evidence, according 
to which the Sokrates of Plato and that of Aristophanes 
l Metn. iv. 6. 13, 


and that of Xenophon are all alike pure fiction, so that 
we really know nothing at all about the man. One of 
the most recent writers on the subject 1 doubts whether 
there is even a grain of truth in the story of the campaigns 
of Sokrates, and denies that he had any relations of any 
kind with Alkibiades. According to him, that was a 
malicious invention of the Sophist Polykrates, 2 who wrote 
a pamphlet against Sokrates before 390 B.C. Plato did 
not stoop to contradict this commonplace pamphleteer, 
and besides, the idea of bringing the two men together 
appealed to him as an interesting one, so he simply wrote 
a romance round it. Now, however incredible such 
theories may appear, they are really far sounder than any- 
thing we can get by picking and choosing whatever we 
please out of Plato, and using it to embroider Xenophon's 
bald tale. It seems to me that we have to choose between 
the Platonic Sokrates and the thoroughgoing nihilism of the 
view just indicated. It is really impossible to preserve 
Xenophon's Sokrates, even if he were worth preserving ; 
and, if we disbelieve the testimony of Plato on the most 
vital points, it is impossible to assign any reason for 
accepting it on others. The Platonic Sokrates would remain , 
indeed, as one of the greatest characters in fiction, but 
some people would find it very hard to read Plato with 
patience, if they supposed him capable of a mystification 
such as this hypothesis credits him with. 

1 A. Gercke in Gercke-Norden, Einlcitung, vol. ii. p. 366 sq. 

2 This statement is based on a passage in the Boustns of Isokrates 
(n. 5), which is supposed to mean that there was not the slightest 
ground for the assertion that Alkibiades was a disciple of Sokrates. As 
I have pointed out (p. 138 n. i) Plato makes Sokrates himself say exactly 
the same thing. It is nowhere suggested in Plato that Alkibiades was a 
fjLOi0y]TYis, or that Sokrates " educated " him. It may be added that the 
Protagoras is almost certainly earlier than the pamphlet of Polykrates, and 
that the relation between Sokrates and Alkibiades is presupposed in it* 



The Associates of Sokrates 

117. We know pretty accurately who composed the 
inner Sokratic circle at the end. In the Phaedo (59 b) 
we have a list of fourteen associates (eraipoi) who were 
present at the death of Sokrates, and to these we must 
add the narrator, Phaidon of Elis, who afterwards founded 
a school of his own. Of these men nine were Athenians, 
Apollodoros, Kritoboulos and his father Kriton, Hermo- 
genes son of Hipponikos, Epigenes, Aischines, Antis- 
thenes, Ktesippos of Paiania, and Menexenos. Xenophon 
also gives us a list of true Sokratics (Mem. i. 2, 48). 
It includes Chairephon, who is absent from Plato's list 
because, as we know from the Afology^ he had died a short 
time before. Kriton and Kritoboulos are also mentioned, 
but not the other Athenians. Apollodoros and Epigenes, 
however, occur in other parts of the Memorabilia^ and 
it is from Hermogenes that Xenophon professes to have 
got his information about the trial of Sokrates. 

The most striking thing about the list, however, is that 
it includes the names of certain foreigners who are known 
to have belonged to Italic schools of philosophy, and who 
are represented as coming to Athens for the express 
purpose of seeing Sokrates before his death. The three 
Thebans, Simmias, Kebes and Phaidondas, were Pytha- 
goreans and disciples of the exiled Philolaos. In the Crito 
(45 b) we learn that Simmias had brought a considerable 


sum of money with him to assist Sokrates in escaping. 
Xenophon also mentions these three in his list of true 
Sokratics, and in another place (iii. n, 17) he lets us 
know that Sokrates had attracted them from Thebes, and 
that they never left his side. In the Phaedo (58 d) the 
Pythagoreans of Phleious are represented as equally en- 
thusiastic. Echekrates says that they are like their guest 
Phaidon in loving above all things to speak of Sokrates 
and to hear about him. Eukleides and Terpsion are 
interesting in a similar way. They were Eleatics and 
lived at Megara. The Academic tradition preserved by 
Cicero makes Eukleides the successor of Parmenides and 
Zeno, and we are told that he " handled " the doctrines of 
Parmenides. The close relation between the Eleatics of 
Megara and Sokrates is further illustrated in the Theaetetus., 
where we are told (143 a) that Eukleides took notes 
of the discourses of Sokrates, and it was with him that 
some of the Sokratics, including Plato, took refuge after 
their Master's death. Besides these men, Aristippos of 
Kyrene and Kleombrotos were expected, but did not 
arrive in time. It is evident that the condemnation of 
Sokrates had' deeply moved all the philosophical schools 
of Hellas. 

1 1 8. Now Plato unquestionably represents the Pytha- 
goreans as sharing a common philosophy with Sokrates, 
and even as looking up to him as its most authoritative 
exponent. It is Sokrates who instructs them in certain 
old doctrines that the contemporary Pythagoreans had 
allowed to drop, and who refutes the theory held both at 
Thebes and Phleious that the soul is an attunement of the 
body. The Eleatic Eukleides is said not only to have 
taken notes of his discourses, but to have had the accuracy of 
these notes confirmed by Sokrates himself when he visited 
Athens. In fact Plato makes all these men regard 
Sokrates as their Master, and it is impossible to suppose 
he could misrepresent their attitude seriously at a time 
when most of them were still living and in close inter- 
course with himself. The suggestion seems to be that, 


after the departure of Philolaos for Italy, Sokrates became 
to all intents and purposes the head of the Pythagoreans 
who remained behind. On one point he is made to 
express surprise that Simmias and Kebes had not been 
instructed by Philolaos (61 d) 5 and Echekrates of Phleious 
is shaken in his belief that the soul is an attunement as 
soon as he is told that Sokrates does not share it (88 d). 
He also accepts the main doctrine of Sokrates as soon as 
he hears it (102 a). 

Plato's account is, I think, confirmed by what we are 
told of Aristoxenos. We know that he was acquainted 
with the last generation of the Pythagoreans at Phleious, 
and that he maintained the doctrine of Philolaos that the 
soul was an Attunement even after he had become a 
follower of Aristotle. We have seen too ( 70) that he 
and his friend Dikaiarchos made a great point of denying 
that Pythagoras had ever practised any of the ascetic 
abstinences and purificatory rites generally attributed to 
him. Now Aristoxenos is the source of a great deal of 
scandalous gossip about Sokrates and Plato. He came 
from Taras and Dikaiarchos from Messene, and Aris- 
toxenos professed to have got his information about 
Sokrates from his father Spintharos, who had known him 
personally. Why should a Tarentine be anxious to 
blacken the character of Sokrates ? The answer suggests 
itself that the friends of Philolaos were annoyed because 
Sokrates had corrupted their doctrine of the nature of the 
soul and had revived the mystical side of Pythagoreanism, 
which they believed they had got rid of once for all 
( 75 75)- ft is at an y rate a fact that they laid special 
stress on the very doctrine of the soul which Plato 
represents Sokrates as refuting. From their point of 
view, he would be just another Hippasos. 


The Forms}- 

119. In the Phaedo the doctrine Sokrates and the 
Pythagoreans are represented as holding in common is 
that of "intelligible forms" (vo^ra e?^), which we have 
seen reason for believing to be Pythagorean in origin 
( 32). Further^ Sokrates is described as making an 
important original contribution to the theory which, in 
fact, completely transforms it. Modern writers generally 
treat this as fiction, and ascribe the doctrine of forms to 
Plato under the name of " the Ideal Theory " or <c the 
Theory of Ideas." The chief ground for this ascription is 
that it is not to be found in the most distinctively Sokratic 
of the dialogues, and it is generally said that it makes its 
first appearance in the Phaedo. That, however, is a circular 
argument ; for the sole ground on which certain dialogues 
have been singled out as specially Sokratic is just that the 
theory in question is not supposed to occur in them. 
There is surely no reason for thinking that Sokrates would 
drag it into all his conversations, and in fact it would have 
been inappropriate for him to refer to it except in talking 
with people who would be likely to understand. Nothing, 
then, can be inferred from his silence on the subject in 
most of the dialogues, especially as that silence is not 
unbroken. By a curious minor epicycle in the argument 
we are warned indeed that, when the doctrine does appear 
to be referred to in a Sokratic dialogue proper, we are not 
to understand the words in the sense they afterwards 
acquired, but this is surely arbitrary in the highest degree. 1 

*I have purposely avoided the word "idea." It inevitably suggests 
to us that the "forms" (tffirj? tSecu) are concepts (vo^/mra), whether our 
own or God's, and this makes a right interpretation of the doctrine 

2 In the Euthyphro, for instance, Sokrates demands that Piety should 
be referred to p.iav rtvot tSeav (5 d), and asks for e/cetvo rb et5os $ vravra 
TO, ocTia ocria, zo-riv (6 c). He also speaks of this as a TrapaSety/ia 
(6 e). In the Meno (72 c) he demands to know the form (etoos) of 
Goodness. In the Cratytus (389 b) we have the highly technical phrase 
avrb o !O*TI /ceyoKt's. I entirely agree with Professor Shorey (Unify oj 


It is much more to the point to observe that the theory of 
forms in the sense in which it is maintained in the Phaedo 
and Republic is wholly absent from what we may fairly 
regard as the most distinctively Platonic of the dialogues, 
those, namely, in which Sokrates is no longer the chief 
speaker. In that sense it is never even mentioned in any 
dialogue later than the Parmenides (in which it is appar- 
ently refuted), with the single exception of the Timatus 
(51 c), where the speaker is a Pythagorean. On the other 
hand, nothing can well be more explicit than the way 
Plato ascribes the doctrine to Sokrates. In the Phaedo it 
is spoken of (100 b) as "nothing new/' but just what 
Sokrates is always talking about. In the Parmenides 
(130 b) Sokrates is asked by the founder of Eleaticism 
whether he had thought of the theory himself, and replies 
in the affirmative. That is supposed to happen at least 
twenty years before Plato was born. Again in the Phaedo 
(76 b), Simmias is made to say that he doubts whether 
ct this time to-morrow," when Sokrates has passed away, 
there will be anyone left who is able to give an adequate 
account of the forms. If that is fiction, it is at least 
deliberate, and I can only ask, as I have asked before, 1 
whether any philosopher ever propounded a new theory of 
his own by representing it as perfectly familiar to a number 
of distinguished living contemporaries some years before 
he had thought of it himself. 

120. The theory which is simply taken for granted in 
the first part of the Phaedo y not only by Simmias and 
Kebes, but also by Echekrates at Phleious, to whom the 
conversation is reported, is as follows. There is a sharp 
distinction between the objects of thought and the objects 
of sense. Only the former can be said to be\ the latter are 
only becoming. It is made clear that the origin of this 

Plato's Though^ Chicago, 1903) in holding that it is futile to look for any 
variation or development of thought in Plato's dialogues down to the 
Repu&fic, though at that point I must part company with him, as will be 

!. Gr. Pt*p. 355- 


theory is to be looked for in the study of mathematics, 
and the distinction between being (ova-La) and becoming 
(yeveo-ii) must be interpreted accordingly. We know what 
we mean by equal, but we have never seen equal sticks 
or stones. The sensible things we call equal are all 
u striving" or "tending" to be such as the equal, but 
they fall far short of it. Still, they are tending towards 
it, and that is why they are said to be becoming. Sensible 
equality is, as it were, equality " in the making " ; but, 
however near it may come to true equality, it never 
reaches it. The connexion of this with the difficulties 
raised by Zeno is obvious. The problem of an indefinite 
approximation which never reaches its goal was that of 
the age. 1 

As we have seen, this theory on its mathematical side is 
essentially Pythagorean. Where it differs from anything 
we can reasonably attribute to the Pythagoreans is in the 
systematic inclusion of what we should call moral and 
aesthetic forms on an equality with the mathematical. 
We have never seen anything that is "just beautiful" 
(ai5ro o <TTL /caXoj/) or "just good " (avro o GCTTLV ayaQov) 
any more than we have seen anything "just equal" (avro 
TO ?CTOJ/). This tends to emphasise that aspect of the 
forms in which they are regarded as patterns or exemplars 
(irapaSeiyftaTa), the " upper limits" to which the manifold 
and imperfect things of sense tend to approximate as far 
as possible. It may sound a little strange to say that an 
isosceles right-angled triangle would be a triangular 
number if it could, but such a way of speaking becomes 
quite natural when we introduce moral and aesthetic 
forms. This is what Aristotle appears to mean when he 
makes the preoccupation of Sokrates with ethical matters 
play so important a part in the development of the theory. 
The Pythagoreans, he tells us, had only determined a few 
things numerically, such as opportunity, justice, and 
marriage, and they had been influenced by superficial 

1 We may illustrate the relation of yevco-i? to ova-ia by the evaluation 
of *r to any number of decimal places. 


analogies ; * it was Sokrates that suggested a systematic 
search for the universal in other fields than mathematics. 2 
It will be observed further that we do not hear in the 
Phaedo of any attempt to connect the forms with numbers, 
and this suggests that the persons whom Aristotle refers 
to as those "who first said there were forms," and dis- 
tinguishes from Plato on that very ground, 3 are no other 
than the persons who call themselves "we" in the Phaedo. 
I do not, however, quote that as external evidence ; for I 
think we shall see reason to believe that everything Aris- 
totle tells us about Sokrates comes from the Platonic 
dialogues, and especially from the Phaedo itself. 4 

121. The account given by Sokrates in the Phaedo of 
the process by which we come to know the forms is apt to 
be insufficiently appreciated because it is expressed in the 
mythical language of the doctrine of Reminiscence, which we 
are expressly warned in the Meno (86 b, 6) not to take too 
literally. The question really is, how we come to have a 
standard which enables us to pronounce the things of sense 
to be imperfect. We certainly do not start with such a 
standard in our possession ; it is only our experience of 
sensible things that gives rise to our apprehension of it. 
On the other hand, our apprehension of the standard 
when it does arise cannot be produced by the sensible 
things, since it is something that goes beyond any or all 
of them. Now when we apprehend a thing, and this 

l Met. M. 3. 1078 b, 21 ; A. 5. 9873, 22. 

2 Afr/. A. 6. 987 b, i. *Met. M. 4. 1078 b, n. 

4 It must be remembered that Sokrates had been dead for over thirty 
years when Aristotle first came to Athens at the age of eighteen. His 
summary and highly ambiguous statements must, therefore, be inter- 
preted, if possible, in the light of the other evidence. To use them for 
the purpose of rebutting it appears to me methodically indefensible. 
That is to employ hearsay and inference to discredit first-hand testimony, 
and we must have some rules of evidence in historical as well as in 
judicial inquiries. I believe that, if we allow for Aristotle's personal 
way of looking at things, his statements can be interpreted so as not to 
do violence to the record ; but, if not, that is a question which concerns 
the interpreter of Aristotle, not the interpreter of Sokrates. 


apprehension gives rise at the same time to the thought of 
another thing which the first thing is either like or unlike, 
we call that being " reminded" or put in mind of the one 
thing by the other (73 c). The sticks and stones we call 
equal are like the equal, and those we call unequal are 
unlike it, but both alike give rise to the thought of what 
is "just equal" (avro TO iVoi/). It follows that, as we 
are put in mind of it both by things that are like it and 
things that are unlike it, our knowledge of the equal must 
be independent of sense altogether. And the same is true 
of "the beautiful itself" and " the good itself. 1 ' 

Aristotle expresses this in his own way by saying there 
are two things that may fairly be attributed to Sokrates, 
universal definitions and inductive reasoning. In the Prior 
Analytics (67 a, 2 1) he definitely associates the doctrine of the 
Meno that learning is Reminiscence with what he calls the 
" recognition" of the universal in a particular case. "In 
no case," he says, " do we find that we have a previous 
knowledge of the particulars, but we get the knowledge of 
the particulars in the process of induction by recognising 
them as it were (<So-xe^> avayvcopHZovras)" There is no 
doubt, then, what Aristotle means by saying that Sokrates 
may be credited with the introduction of inductive reason- 
ings, and it is exactly the process described in the Phaedo. 
It is also correct to say, as he does, that the universal 
which we come to recognise in this way is " the What is 
it?" (TO rt <TTL) ; for in the Phaedo (78 d) Sokrates 
describes the sort of reality possessed by the forms as 
" that of the being of which we give an account in our 
questions and answers," that is, in the dialectic process. It 
will be observed that there is nothing here about abstract- 
ing the common attributes of a class and setting it up as a 
class-concept. That is a modern gloss on Aristotle's words, 
and his reference to the Meno shows he was quite aware of 
the real meaning of the doctrine of Reminiscence. There 
is nothing to suggest, then, that what he says on this point 
is derived from any other source than Plato's dialogues. 
He has expressed the thing in his own way, no doubt, and 


it may be a question whether it does full justice to the 
doctrine of Sokrates, but that is another matter. If he 
was to express it in his own language, he could hardly say 
anything else, and, after all, his own theory of induction is 
much more like the doctrine of Reminiscence than the 
travesty of it given in some text-books. It should be 
added that, when Aristotle says certain things may " fairly" 
(SiKalw} be attributed to Sokrates, he is thinking, as he 
often does, of earlier philosophers as contributing certain 
elements to his own system, and that he is contrasting 
Sokrates in this respect with the Pythagoreans. He is not 
thinking of any distinction between the "historical" and 
the "Platonic" Sokrates, and there is no evidence that he 
ever made such a distinction. 

122. Now it is with the soul by means of reasoning 
(Xo^ttTAio?) that we apprehend the forms, while particulars 
are apprehended through the body by sensation. Indeed, 
the body and its senses are only a hindrance to the acquisi- 
tion of true wisdom, and the more we can make ourselves 
independent of them, the nearer we shall come to the 
knowledge of reality and truth. We have seen that the 
things of sense cannot be said to have being (ova-la) at all, 
but only becoming (yeWi?), and that they are merely like- 
nesses or images of the eternal and immutable standards 
or patterns (TrapaSelyftaTa) we are forced to postulate. Of 
these alone can there be knowledge ; our apprehension of 
the things of sense is only "imagination" (eiW/a) 1 or at 
best belief (&>', -jr/crrt?). If we would have true know- 
ledge, we must seek to rid ourselves of the body, so far as 
that is possible in this life ; for it is only when the soul has 
departed from the body that it can have knowledge in its 
purity. Yet even in this life, by the practice of dying 
daily, we may so far mortify the flesh that for a brief space 
we may behold the eternal realities in a vision, and so 
being "out of the body" obtain a foretaste of immortality. 

l Rep. 5 34 a. There is an untranslatable play on words here; for 
etKaxria, is properly "guess work" (from /ca(<r0cu), but it also suggests 
the apprehension of images ( 


Such is the teaching of the first part of the Phaedo, and 
there can be no doubt that it points to an almost com- 
plete severance of the world of sense from the world of 

123. But then, by one of those dramatic surprises so 
characteristic of Plato's dialogues, when we have been 
raised to this pitch of spiritual elevation, we are brought 
to the ground once more, and made to feel that, however 
beautiful and edifying the doctrine may be, it does not 
really satisfy us. It is Plato's way to mark the importance 
of the different sections of an argument by the length and 
elaboration of the digressions that precede them. In the 
present case he uses every resource of his art to make us 
feel that we are coming to something fundamental. In 
the first place, there is a long and ominous silence (84 c), 
broken at length by a whispered conversation between 
Simmias and Kebes. Sokrates sees they are not convinced, 
and he urges them to state their difficulties; for, as he 
allows, the doctrine is open to many objections if we discuss 
it seriously. Then follows (846) the magnificent passage 
in which he compares himself to the dying swan who sings 
in praise of their common master Apollo, the lord of 
Delphoi and of Delos, who had played so mysterious a 
part in the life of Sokrates himself, and was also the chief 

fod of the Pythagoreans. Simmias replies (850) that 
okrates no doubt feels with him that certain knowledge is 
impossible on such subjects, but that we must test and try 
all theories 3 and, in default of some divine doctrine ($eZo? 
Ao'-yo?), make the best of the human one that approves 
itself most. The particular difficulty he feels is just the 
theory, of which we have seen the great historical impor- 
tance, that the soul is an attunement (apjuovla) of the body, 
and cannot therefore be immortal (8 5 e). Kebes has a 
different theory, of which we do not hear elsewhere, but 
which seems to be Herakleitean in origin, namely, that the 
soul is the organising principle of the body which it 
weaves as a garment. The body is always being worn out 
and woven afresh, and thus the soul may properly be said 

FENE2IS KAI $60PA 161 

to outlast many bodies. That does not prove, however, 
that one of these bodies may not be the last, and that the 
soul may not perish before it (88 b). We are told (88 c) 
that the effect of these words was to produce a feeling of 
profound dejection In the company. They felt as if they 
could never trust themselves to believe any doctrine again, 
since this one had been so easily overthrown. The narra- 
tive is even interrupted, and we are taken back to Phleious, 
where Echekrates says the same effect has been produced 
on him. Then comes the warning of Sokrates against 
Cl misology," or hatred of theories. It is just like misan- 
thropy, which arises from ignorance of the art of dealing 
with men. Just as the man who knows the world knows 
that very good men and very bad men are equally rare, so 
the man who knows the art of dealing with theories will 
not expect too much of philosophical doctrines, but neither 
will he lose faith (89 d 5^.). The impression intended to 
be left on us by all these digressions Is certainly that the 
doctrine of forms as expounded in the earlier part of the 
dialogue Is somehow inadequate, and we are prepared to 
find that it will be considerably modified in the sequel. 
We are also intended to understand that the later Pytha- 
gorean view of the soul is a serious obstacle to a sound 

124. This doctrine is disposed of without much diffi- 
culty, chiefly by the consideration that, if the soul is anattune- 
ment and goodness is an attunement, we have to assume an 
attunement of an attunement,so that one tuning will be more 
tuned than another. The theory of Kebes, however, raises 
a far more fundamental question, namely, that of the cause 
of coming Into being and ceasing to be (yevea-ts teal <p6opa). 
To say that becoming is an image or likeness of being 
explains nothing at all. It really amounts to saying that 
there Is a world of sense which is a vain show, standing in 
no intelligible relation to reality. Unless we can overcome 
this separation between appearance and reality in some way, 
we cannot say anything at all, and least of all that the soul 
is immortal. What we want is not merely a theory of 


being (ouo-m), but also a theory of becoming 
It is at this point that Sokrates gives the sketch of his 
intellectual development already referred to ( 103); and, 
if words mean anything, it must be implied that we are 
now coming to his personal contribution to the doctrine. 
He speaks of this (97 b, lood) with characteristic irony 
as a "silly and muddled" theory, and calls it a makeshift 
or pis-caller (Sevrepo? TrAoi/?, 99 d), but we must not be 
deceived by this way of speaking. It is also the hypo- 
thesis from which he will not suffer himself to be 
dislodged by anyone, and he believes it to be capable 
of showing the cause of coming into being and ceasing 
to be in the world of sensible experience, a thing the 
earlier form of the doctrine could give no intelligible 
account of. 

125. Sokrates tells us, then, that when he could find 
no satisfaction in the science of his time, and in particular 
no answer to the question of the cause of becoming and 
ceasing to be (yevco-is KCU <fr6opa)> he resolved to adopt a 
new method of inquiry. He would no longer consider 
the question from the point of view of the things (ev roZ? 
Ijoyot?) but from that of the judgements we make about 
them and the propositions in which these are expressed 
(ev TOI? \6yoi9). He is represented both in the Meno and in 
the Phaedo as much impressed by the efficacy of the mathe- 
maticians' method of " hypothesis," which Zeno had made 
matter of common knowledge at Athens by this time. 
To understand its meaning, we must leave out of account 
for the present the special use of the term " hypothesis " 
in Aristotelian Logic, and also the popular etymology 
alluded to by Plato in the Republic (511 b) which regards 
the primary meaning of the word as foundation or basis, 
a sense in which it is not used. If we do this, we shall 
be struck at once by the fact that the corresponding verb 
(yiroriQevQai) has two chief significations, firstly that of 
setting before oneself or others a task to be done, and 
secondly that of setting before oneself or others a subject 
to be treated, in a speech, for instance, or a drama. This 


usage is as old as Homer, 1 and by a natural extension the 
verb is freely used in Ionic of suggesting a course of 
action. That way of speaking accounts for Euclid's use 
of the word "given," and also of perfect imperatives like 
" let there be given " (SeSo<r6a>). The original idea is that 
of a piece of work given out to be done, and the proposi- 
tion accordingly ends up with a statement that it has been 
done (Q.E.F. o-rrep ei 7rovj<rai or Q.E.D. OTrep e^et Sei^ai). 
The procedure is as follows. It is assumed that the 
proposition stated in the "hypothesis" is true (or that 
the required construction has been performed), and then 
the consequences (ra crvju,(3aivovTa) of that assumption are 
deduced till we come to a proposition we know to be true 
(or a construction we are able to perform). If, however, 
we come to a proposition which is absurd (or to a con- 
struction which is impossible), the hypothesis is " de- 
stroyed " (avaipeiTaL) tollitur). The regular terminology 
accordingly is, " if A is B, what must follow ? " (rl ^pr; 
crvjuLfiaLveiv ;), and that explains why the conjunction "if" 
has come to be regarded as the mark of a hypothesis, 
Plato's Parmenides is the locus classicus for all this, but the 
method is older. In the Hippokratean treatise on Ancient 
Medicine^ the fundamental doctrines of Empedokles and 
others are called hypotheses, and the key to this way of 
speaking is also to be found in Plato's Parmenides. There 
the doctrine of Parmenides is referred to as the hypothesis 
If it is one, and that of his opponents as the hypothesis 
If there are many? In the same way the hypothesis of 
Empedokles might be stated in the form If there are four. 
This is a result of the Eleatic dialectic. It is not implied 
in the least that Parmenides or Empedokles regarded their 
theories as " merely hypothetical." That is a far later 

1 See Liddell and Scott, s.v. wrorifty/u, ii. 2. The materials for a 
correct account of the term vTro&crts are also to be found in Liddell and 
Scott, s.v., but they require rearrangement. The article should be read 
in the order iii., iv., i, 2 y ii. 2, ii. I. 

2 Farm. 1 28 d, 5. The reading of the best MSS. and Proclus is 
&jr6$Tis et iroXXd Icrrw. 


use of the word. It is only meant that their method of 
exposition was to trace out the consequences of their 
fundamental postulates. We can see for ourselves that 
this is what Parmenides does in his poem. Zeno syste- 
matised the procedure, and it was doubtless from Zeno 
Sokrates learnt it 

Like all dialectical methods, this procedure is subject to 
strict rules. We first take a statement which appears to 
have a high degree of probability, and we set down as true 
whatever agrees with that and as false whatever does not. 
It is not allowable for the answerer to raise any questions 
about the hypothesis itself till this has been done, and until 
it is seen whether the consequences of the hypothesis 
involve anything absurd. If they do not, and there is 
still any doubt about the hypothesis, the answerer may 
question it, but not till then. The deduction of conse- 
quences must be kept quite separate from the question of 
the truth of the hypothesis. If that is not admitted even 
then, we may go on to show that it is a consequence of 
some higher hypothesis which we assume in the same way, 
till at last we come to some hypothesis which is adequate 
in the sense that the answerer accepts it (101 d). It will 
be seen that there is no question of demonstrating this 
ultimate hypothesis ; it only holds good because it is 
accepted by the other party to the discussion. The whole 
fabric depends on the agreement of the two parties to the 

126, In the present case, the hypothesis Sokrates starts 
from is the distinction of the sensible from the intelligible, 
which is of course allowed to be true by his Pythagorean 
interlocutor without any hesitation (looc). Assuming, 
then, that there is a form of the beautiful, we have next 
to ask what makes us call a particular thing beautiful. It 
is no answer to say it has a bright colour or anything else 
of the kind ; that throws no light on the meaning of the 
statement, " This is beautiful." On the one hand, this is, 
of course, the problem of predication, the question of what 
is involved in saying " A is B," but that is not quite the 


form it takes in the Phaedo. We are discussing coming 
into being and ceasing to be (yei/ecro KCU <p6opa), or, in 
other words, we are asking how there can be a world of 
becoming alongside of the world of being which alone is 
the object of knowledge. The question is better formu- 
lated, then, if we say "What makes a thing beautiful?" 
The " simple-minded answer " Sokrates gives to this is : 
If there is anything 'beautiful besides Beauty itself , Beauty makes 
it beautiful, and this is explained to mean that it is the 
" presence " (irapovo-ld) of the form in it that makes any- 
thing beautiful or whatever else we say it is. The pre- 
dicate of a proposition is always a form, and a particular 
sensible thing is nothing else but the common meeting- 
place of s. number of predicates, each of which is an 
intelligible form, and in that sense there is no longer a 
separation between the world of thought and the world 
of sense. On the other hand, none of the forms we 
predicate of a thing is present in it completely, and this 
relation is expressed by saying that the thing cc partakes 
in " the forms that are present in it. Apart from these, 
it has no independent reality ; and, if we know all the 
forms in which anything participates, there is nothing 
more to know about it. The doctrine is most distinctly 
stated in the Republic (476 a), where we are told that each of 
the forms is one, but by reason of their communion (KOIVWICL) 
with actions and bodies and with one another, they appear 
everywhere, and each seems to be many. 1 It is in that 
sense that Sokrates the Sokrates of the Phaedo and the 
Republic does not separate the forms from the world of 
sensible particulars, 2 and it is just because he denies all 
reality to the sensible particulars except what they derive 
from the partial presence of the forms in them. The 

1 The /coivowa of the forms with one another in the sensible world is 
quite different from their Kotvcovta with one another in the intelligible 
world which Plato taught. That is just where Plato differs from Sokrates, 
as we shall see. 

2 Ar. Met. M. 4. zoySb, 30. dAA' o p*v *2wKpdrr}<s rot KaBoXov oi 


Pythagorean doctrine of imitation left the sensible and 
intelligible as two separate worlds ; the doctrine of partici- 
pation makes the sensible identical with the intelligible, 
except that in sensible things the forms appear to us as a 
manifold instead of in their unity, and that they are only 
imperfectly embodied in the particulars. We should not 
be entitled to predicate the form of the thing unless the 
form were really in it. 

127. We may say, then, that the problem of Sokrates 
was to show how it was possible for the things of sense 
to be real, and he answers it by saying that they are 
real in so far as they partake in reality or as reality is 
present in them. He is conscious that these are meta- 
phorical expressions, and so is the formula he substitutes 
in the latter part of the dialogue, namely, that the form 
"occupies" or "takes possession of" (rar^ei) particular 
things. That way of putting the matter is adopted in the 
course of the final argument for the immortality of the 
soul, which, though not an object of sense, is nevertheless 
a particular thing and not a form. The proof is briefly 
that, from its very nature, the soul partakes in the form 
of life or is "occupied" by it, and it is shown that a thing 
which is necessarily and of its own nature occupied by 
a given form will not admit the form opposite to that. 
If attacked by it it will either withdraw or perish. The 
soul cannot perish, however, so it will necessarily with- 
draw. For reasons which will be obvious, Sokrates him- 
self is not altogether satisfied with this argument, and 
Plato found it necessary to defend the belief in immor- 
tality in quite another way. The real result of the Phaedo 
is not^this, but simply that no particular thing can become 
anything except by partaking in, or being occupied by, 
the form of what it becomes, nor cease to be anything 
except by ceasing to partake in the form. 1 Such is 
the doctrine Plato attributed to Sokrates, and it is as 

1 This is how Aristotle formulates the theory of the Phaedo in Gen. 
Cotr. B. 6. 335 b, 10. He does not attribute it to Plato, but to 
"Sokrates in the Phaedo? 


clearly distinguished from his own as from that of the 

128. But though the Pythagorean separation (^apia-fioi) 
of the things of sense from the things of thought has been 
overcome, it still remains true that there is a gulf between 
the confused manifold of sense and what is called in 
the Phaedrus (247 c) the " colourless, shapeless, intangible 
reality" beheld by thought alone. This gulf the soul is 
ever seeking to bridge over, and its striving can only be 
described in the language of passionate love. That is 
involved in the very name of philosophy itself, and is 
brought home to us by calling philosophers " lovers of 
wisdom " (epaa-Toi <ppovYj(reu>i), where the verbal variation 
is meant to remind us of the original meaning of the 
name. No one who is wholly dull and stupid feels this 
craving, nor does he who is already wise, as God is. 
Love is the child of Poverty and Resource. Now the soul 
itself and its strivings can only be adequately described in 
mythical language ; for they belong to the middle region 
which is not yet wholly intelligible. The objects of its 
yearning are not mythical at all. The inspired lover is 
seeking the intelligible just as much and more than the 
mathematician, and I can see no ground for holding that 
even in the Phaedrus^ the forms are regarded as super- 
natural "things" of any kind. The "supercelestial region" 
is clearly identified with that of pure thought, and the 
forms the mind beholds in it Righteousness itself, 
Soberness itself, Knowledge itself do not lend them- 
selves in any way to crude pictorial fancies. It is true 
that our relation to this supreme reality can only be 
expressed in the language of feeling, but it is not by 
feeling we apprehend it when and in so far as we can 
do so. It is expressly said to be visible to mind alone 
(papa Qearrri v]. There is no suggestion of a different 
way of knowing to which we may have recourse when 
reason and intelligence fail us. To put the matter in 
another way, allegory and myth are not employed to 
express something above reason, but to adumbrate what is 


below reason, so far as that can be done at all. It has its 
place half-way up the scale and not at the top ; for it 
is only the poverty Love inherits from his mother that 
gives rise to these passionate yearnings. When they are 
satisfied, there is no more room for striving and longing. 
I suspect that all true mysticism is of this nature, and 
that to set feeling above reason as a means of krowing 
is only a perversion of it. However that may be, 
I am firmly convinced that the mystical side of the 
doctrine of forms is due to Sokrates and not to Plato. 
We know certain facts about him, such as his " voice " 
and his trances, which prove him to have possessed the 
mystic temperament, and we know certain facts which 
explain the manner in which he conceives the mystic 
love. On the other hand, we have seen that there was 
another side to his nature which would safeguard him 
from the spurious kind of mysticism. I entirely agree 
with the demand 1 for a psychological explanation of the 
two sides of the doctrine of forms, but the soul in which 
that is most easily to be found appears to me to be the 
soul of Sokrates, son of Sophroniskos. It is certainly in 
the Symposium that we have the most vivid picture of his 
personality, and there the " enthusiasm " and the " irony " 
are in perfect unison. 

129. Nevertheless the Sokrates of the Phaedo does not 
succeed in reaching the goal he has set before himself. 
He had turned away from the science of his time just 
because it could not show how everything is as it is 
because it is best for it to be so ; and, though coming 
into being and ceasing to be have been explained in a 
sense, we cannot be said to be much nearer the fulfilment 
of that demand. That is because we have assumed certain 
forms which serve to explain the world of experience ; 
but we have not gone on to examine this hypothesis itself 

1 See Professor Stewart's Myths ofPlato, which is far the best treatment 
of this part of the subject. It will be obvious that I am obliged to 
differ from it in some important respects, but that does not impair my 
appreciation of the work. 


in the light of a higher one, and therefore we cannot 
say why there should be a world of experience at all. 
Sokrates is represented as quite conscious of this in the 
Republic. There he is made to say (505 d sqq^ that 
we must look at all the other forms in the light of the 
Form of the Good, which is no mere hypothesis, but 
the true starting-point of knowledge. He confesses 3 how- 
ever, that he can only describe it in a parable, and it is 
never referred to again in Plato's dialogues. The passage in 
the Republic stands quite by itself. We can see dimly what 
the Good must be if we liken it to the Sun, which is the 
cause both of growth and of vision in the sensible world, 
though it is neither growth nor vision itself. In the same 
way the Good must be the cause of knowledge and being 
in the intelligible world, though it is neither of these, but 
far beyond both of them in glory and power. 1 It is very 
significant that Sokrates is made to regard this purely 
negative characterisation of the Good as marking a failure 
to apprehend its true nature ; it was left for thinkers of 
a later age to find satisfaction in it as a positive doctrine. 
That Sokrates really did speak of it in some such way 
as this appears to be proved by the fact that Eukleides of 
Megara identified the Good with the Eleatic One. That 
seems to be how he reconciled his Eleaticism with his 
position as an "associate" of Sokrates. The Pytha- 
goreans would have little or no difficulty in accepting 
the doctrine of the Phaedo^ but an Eleatic could not be 
expected to acquiesce in a plurality of forms. If Sokrates 
hinted at the ultimate unity of all the forms in the Good, 
we can understand what Eukleides meant ; otherwise it 
would be very hard to follow him. Even so, there is 

1 This language has led some to identify the form of the Good with 
God, but that is certainly wrong. God is a soul and not a form, and in 
the Timaeus (which, as we shall see, represents a highly developed form 
of Pythagoreanism) the Good is above God. The difficulties raised 
by this doctrine led in later days to the conception of a highest and 
unknowable God and a secondary creative God (the Demiurge), but 
there is no trace of this till Hellenistic times. The Demiurge of the 
Timaeus is the highest God there is. 


a rift here in the doctrine of the Sokratic society, and 
we shall see how important that became in the next 


130. The theory of goodness Plato attributes to 
Sokrates is only intelligible in the light of the theory 
of knowledge and reality we have been considering. It is 
made clear, in the first place, that he was led to formulate 
it because he was dissatisfied with the teaching of the 
"Sophists," and we must try to understand exactly where 
he differed from them. No doctrine is more closely 
associated with the name of Sokrates or better attested 
than that of the identity of goodness and knowledge, with 
its corollary that no one is voluntarily bad. No one who 
really knows what is good and what is bad can possibly 
choose the bad, and badness is, therefore, in the last 
resort, a form of ignorance. That Sokrates held this 
doctrine is more universally admitted than any other fact 
whatsoever about him. 

That being so, it is not a little remarkable that, in a 
considerable number of his dialogues, Plato represents 
Sokrates as arguing against the doctrine, at least in its 
most obvious sense. He is made to say, for instance, 
that goodness cannot be knowledge ; for, if it were, the 
great statesmen of Athens would certainly have taught 
their own goodness to their sons, whereas most of these 
were complete failures. Nor can it be said that the 
" Sophists" really teach it; for then these same states- 
men would have had their sons taught goodness just as 
they had them taught riding and music. In fact, goodness 
appears to be something that comes by chance or divine 
favour (0e/a nolpa) to some people and not to others. 
Those who have it can give no account of it ; they cannot 
even tell what it is, and are therefore quite unable to 
impart it. They are like the poets who compose under 
the influence of inspiration of some kind, and cannot even 
give an intelligent interpretation of their own works. 


The connexion of this with what we are told about the 
mission of Sokrates in the Apology is obvious. 

Nevertheless, the contradiction between these statements 
and the doctrine that goodness is knowledge is puzzling 
at first sight. It has been said, of course, that in these 
dialogues Plato is feeling his way to a more advanced 
doctrine than that of " the historical Sokrates," but this 
line of interpretation breaks down as usual. It is perfectly 
certain that the arguments about statesmen and their 
sons was actually used by Sokrates himself, and we gather 
from the Meno and from Xenophon that it was one of the 
things that annoyed Anytos. As for Plato, he still main- 
tains the doctrine that goodness is knowledge, and that no 
one is voluntarily bad, in his very latest work, the Laws 
(860 d). 

131. It will help us to understand this difficulty if we 
remember that the identification of goodness and know- 
ledge was not really a doctrine peculiar to Sokrates, but 
was implied in the general belief of his time that goodness 
could be taught. The question between Sokrates and his 
contemporaries was not that, but the much more funda- 
mental one of what goodness was identical with knowledge 
and therefore teachable. The Sophists were not wrong in 
holding that goodness could be taught ; they were wrong 
in so far as the goodness they professed to teach was just 
that which, not being knowledge, could not be taught, and 
in so far as they ignored altogether that higher kind of 
goodness which alone was knowledge and therefore alone 
teachable. If we attribute this distinction to Sokrates we 
shall find no real contradictions in the dialogues dealing 
with the subject. 

Nor are we without external evidence in support of this 
view. In the Helen of Isokrates (10. i) we read that there 
are certain people who pride themselves on setting up 
a paradox and arguing tolerably in favour of it. "Some 
have grown old in denying that it is possible to say what 
is false, or to contradict, or to make two opposite state- 
ments about the same thing." That, no doubt, is meant 


for Antisthenes. cc Others argue in detail that justice, and 
courage and wisdom are the same thing, and deny that 
any of these things come by nature, saying that there 
is one knowledge of them all." That, I take it, refers to 
Sokrates. c< Lastly, there are those who spend their time 
in contentions (irepl ra? eplSas SiaTpl/Sova-i)." Plato uses 
that phrase too, and we shall have to discuss its application 
later. A little further on (10. 5) Isokrates makes light 
of the distinction between knowledge (eTHcrr^?;) and belief 
(c?oa), asserting that it is better to have a reasonable belief 
about useful things than a precise knowledge of what is 
useless. Similarly in his pamphlet Against the Sophists, 
he speaks (13. i) of those who spend their time in disputa- 
tions, and who profess to teach the young their duties and 
how to attain happiness (13. 3). Here too knowledge 
and belief are contrasted, and finally Isokrates denies that 
righteousness and morality can be learnt. 

It is very difficult to believe that any of these references 
can be intended for Plato, as is often supposed. Isokrates 
was older than Plato, and both the Helen and the tract 
Against the Sophists are dated with probability some time 
before 390 B.C., when Isokrates opened his school, and 
therefore some time before Plato came forward as a 
teacher. It is plain too that Isokrates is concerned with 
the educational theories of his immediate predecessors, 
and it is not very likely he should go out of his way 
to attack a younger contemporary whom he had no reason 
at that date to regard as a rival. On the other hand, the 
question of Sokrates was very actual indeed at the time ; 
for^the Sophist Polykrates had just published his pamphlet 
against him, with the object of showing he was rightly put 
to death for the bad influence of the education he gave. 
We know too from the Bousiris that Isokrates had busied 
himself with this pamphlet. He must, then, have wished 
to make his attitude to Sokrates quite clear, while there 
was no reason for him to trouble about Plato yet awhile. 
But, if that is so, we may safely attribute the distinction 
between belief (Sdga) and knowledge (emcrntfui) to Sokrates 


himself, and also the doctrine that goodness is one and 
that the knowledge of it is one, and that means in turn 
that there is no difficulty in attributing to Sokrates himself 
the whole theory of goodness expounded in Plato's earlier 
dialogues down to and including the Meno, and even, 
in substance, that set forth in the Republic. 

132. We are left in no doubt as to what "goodness" 
(apeTTJi) meant in the language of the time. The Sophists, 
we have seen, professed to teach the goodness of the man 
and the citizen, and that was explained as the art of manag- 
ing states and families rightly. It was, in fact, what we 
call efficiency. To the Greeks goodness was always some- 
thing positive ; it meant a habit of soul that enabled 
the possessor of it to do something, and not merely, as it 
is apt to mean with us, one that leads him to abstain from 
doing any particular harm. No Greek would have called 
a man good on purely negative grounds like that ; he 
must be good for something. So far neither Sokrates nor 
Plato nor Aristotle would have the least quarrel with 
the current view. We have seen, however ( 88), that 
the political condition of Athens was such in those days 
that the word tended to acquire a peculiar colour. That 
comes out better than anywhere else in the passage of 
Thucydides where he tells us that Antiphon, the chief 
contriver of the Revolution of the Four Hundred, was 
second to no other Athenian in " goodness " (aperrf). That 
was, in practice, the only sort of goodness the Sophists 
had the opportunity of teaching ; for it was the only sort 
demanded by those who could pay for it. It amounted 
to little more than skill in the arts of party intrigue. 

The goodness Sokrates identified with knowledge was 
naturally of a different order, but he always admitted 
the relative value of " true belief" (akyOw &>'a) for 
practical purposes. In the Meno he says (97 b) that 
if you want to go to Larissa a true belief about the way 
will take you there as well as knowledge. There is noth- 
ing astonishing in such an admission in view of the 
account we have given of his theory of knowledge. As 


we have seen, he was very far from denying the relative 
value of ordinary experience. Its objects are the same as 
those of knowledge, though they are imperfect and con- 
fused. He never meant to say that the great states- 
men of Athens did no good at all, or to deny all value 
to the works of the poets. If the statesmen of the past 
had no goodness of their own, there would be nothing 
surprising in their failure to impart goodness to their sons. 
The weak point of such goodness, however, is that it 
is not based on any rational ground (Xo'yo?) and cannot 
therefore be counted on. It is mainly an affair of tempera- 
ment and happy chance. It is only, we are told in the 
Meno (98 a), when it has been chained fast by a reasoned 
knowledge of its cause (ama? Xoy/o-^a!) that we can be 
sure of its not running away like the Statues of Daidalos. 
Then, and then only, do we have goodness which is also 
knowledge and can therefore be taught. 

It will be observed that this theory of goodness and the 
good is the exact counterpart of the theory of knowledge 
and reality which Plato ascribes to Sokrates, and this is 
another indication of the correctness of that ascription. 
Just as we cannot explain the cause (ama) of things in the 
world of coming into being or ceasing to be unless we 
regard them as participating or ceasing to participate in an 
intelligible "form," so we cannot have true goodness 
unless each act is referred by reasoning (XoyjayxoY) to its 
true cause (ama). Everyday goodness is just like the 
world of sensible experience in that it is inconstant and 
variable ; true goodness must be constant and invariable. 
According to the Phaedo (82 a) Sokrates distinguished the 
two as "philosophic goodness" (<pi\o<ro<piKrj apeTrf) and 
"popular goodness" (SwoTuc)] aperrj}^ or the "goodness 
of the citizen" (TroXm/4 apeTri). The former depends on 
intellect (j/ovj), the latter on habit (edos). It is the former 
alone that is teachable ; for it alone is knowledge, and 
nothing can be taught but knowledge. The latter is only 
good at all in so far as it participates in the former. Apart 
from that, it is a shifting and uncertain thing. 


133. But though goodness in the full sense of the 
word is knowledge, it is not an art, that is to say, an 
external accomplishment that may be acquired by anyone, 
and which he may exercise or not at his pleasure. Plato 
has given us at full length two very similar arguments on 
this point, and they bear every mark of being genuinely 
Sokratic. In particular their constant reference to the 
practice of artificers is highly characteristic. The best 
known is the argument with Polemarchos in the Republic, 
which is less likely to be misunderstood if read in the light 
of the other, which occurs in the Hippias minor. In the 
Republic (332 e sqq.} the argument is directed to showing 
that, if goodness is an art (a view for which Polemarchos 
and not Sokrates is responsible) the honest man will be 
the best thief, just as the doctor will be the most successful 
murderer. The argument of the Hippias minor is that 
wisdom is required as much or more to tell lies as to tell 
the truth, and that it is better to do wrong voluntarily 
than involuntarily. The point is the same in both cases. 
An art or capacity (Svva/jni) is always " of opposites." The 
man who can make a good use of it is also the man who 
can make a bad one, and therefore something more must 
be implied in goodness than this. That too was forced on 
Sokrates by the practice of the Sophists. Gorgias disclaims 
all responsibility for the use his pupils may make of the 
art of Rhetoric which they learn from him. We have no 
more right, he says (456 d) to blame the teacher of rhetoric 
for the misdeeds of his pupils than we should have to 
blame the teacher of boxing if his pupil used his art to 
injure his neighbours. The question involved in the 
argument with Polemarchos is really the same. Is it 
possible to regard goodness as a purely neutral accomplish- 
ment of this kind, or is it something that belongs to the 
very nature of the soul that possesses it, so that it is 
really impossible for the good man to do evil or to injure 
anyone ? 

134. Another question that was much discussed at 
this time was that of the unity of goodness, and to Sokrates 


this question was closely bound up with the other. The 
professional teaching of goodness was apt to suggest that you 
could learn one branch of it and not another. You might, 
for instance, learn courage without learning honesty, or 
vice versa. If the different forms of goodness are so many 
" arts " or external accomplishments, they will stand in no 
necessary connexion with one another, and we cannot 
say that goodness is one. Sokrates approaches this 
question from the point of view of the different kinds of 
goodness. The Laches, for example, starts from courage, 
and the Charmides from soberness. In both cases the 
particular virtue under discussion is identified with know- 
ledge, but the identification is not made by Sokrates. On 
the contrary, his argument is entirely directed to showing 
that, if we identify any particular form of goodness with 
knowledge, it is impossible to maintain any distinction 
between it and any other form of goodness. From that 
point of view they all become merged in one. 

Both these doctrines, that of the unity of goodness, and 
that which refuses to identify goodness with an art, are 
supported by another line of argument of which Sokrates 
is fond. A good example of this too is to be found in the 
argument with Polemarchos in the Republic (332 c). It is 
that, if you identify any form of goodness with an art, it is 
impossible to discover any use for it. The whole field is 
already covered by the particular arts appropriate to each 
department, and there is no room for the " virtue." One 
might suppose that honesty or justice was a virtue useful 
in partnerships, but we should all prefer a good player to 
an honest man as our partner in a game of skill or as an 
accompanist to our singing. If goodness is looked at in 
this way, it will have no special function to perform ; there 
is no room for it alongside of the other arts. It may be 
harmful, since it is a capacity of opposites, and it is in any 
case superfluous. 

*35- What, then, is the knowledge with which true 
goodness is to be identified ? In the first place it is know- 
ledge of what is good for the human soul. It is at this 


point we see most clearly how the theory of conduct taught 
by Sokrates, like his theory of knowledge, was influenced 
by Pythagorean doctrine. The Pythagoreans had already 
regarded the health of the soul as something analogous to 
the health of the body, and for them this was much more 
than a metaphor. We have seen ( 75) how they arrived 
at their fundamental notion of an attunement (apimovla) or 
blend (/cpacro), and it was this that dominated all medical 
theory so far as it fell under Pythagorean influence. It 
was partly the necessity of explaining goodness in this way 
that made Sokrates reject the later Pythagorean view that 
the soul itself was an attunement ( 124), and he preferred 
to work out the idea from the point of view of what was 
probably an earlier Pythagorean doctrine, that of the parts 
of the soul. In the Gorgias (504 a syy.') Sokrates says that 
goodness is due to the presence of arrangement (TC!^) and 
order (/coV^o?) in the soul, and that this can only be pro- 
duced by knowledge, not by experience or routine. In 
the Republic the same theory is worked out in the most 
elaborate fashion. It is shown that there are three parts 
of the soul, the philosophical or reasoning part ((^AoVo^oi/, 
\ojL<rTLK6v^ temper (^u/zo?), and desire (eiriQufuoL). The 
special virtues of each of these are wisdom, courage, and 
soberness, while justice or righteousness is the principle 
that keeps them all in their proper place. It is shown 
how each of these virtues is represented in the different 
classes of a well-ordered State, and we learn from a con- 
sideration of that how the inner polity of the soul should 
be ordered. We see that wisdom should command, while 
temper assists in the execution of these commands, and 
how the desires should be confined to their proper task of 
supplying the necessary material basis for the rest, and how 
all this is to be secured by justice, which assigns to each 
its proper part and sees that it keeps to it. It is shown 
further how inferior types of State arise from the usurped 
supremacy of one or other of the subaltern parts of the 
soul, and how there are inferior types of character corre- 
sponding to each of these and arising from the same cause. 



No doubt the elaboration of this idea which we find in the 
Republic owes much to the artistic genius of Plato, but it 
appears to me quite certain that the leading idea is Sokratic, 
and indeed Pythagorean. Plato's own view of the soul 
was so different that he would not naturally fall into this 
way of expressing himself, though he might quite well use 
it for purposes of more or less popular exposition. As we 
shall see, it is improbable that he had a definite original 
philosophy of his own by the time the Republic was 
written. 1 

136. This account of the Sokratic philosophy is in 
brief that to which Plato gave currency within fifteen 
years or so of his master's death. It is, I submit, an 
intelligible and consistent whole, and it is quite different 
from anything Aristotle ever ascribes to Plato himself. If 
Plato had originally taught this sytem, and if the doctrine 
Aristotle ascribed to him was only a development of his 
later years, we may be sure that we should have heard 
something about this remarkable change of opinion. As 
it is, there is no hint anywhere in Aristotle that Plato 
ever taught anything else than what he regards as the 
genuine Platonic doctrine. It is impossible, of course, to 
decide the matter finally till we have seen what Plato's 
own philosophy was, but there are two considerations I 
should like to urge before leaving the subject. In the 
first place, it is surely worth while to try the experiment 
of taking Plato's dialogues in their natural sense. That is 
the " hypothesis " on which this work proceeds, and it can 
only be destroyed if we come to consequences that are 
impossible or untrue. In the second place, I would urge 
that the burden of proof does not lie with those who 
adopt this hypothesis, but with those who deny it. We 
cannot be forced to regard the Sokrates of Plato as a 
fiction unless some really cogent argument can be produced 
for doing so, and I am not aware that this has ever been 

1 1 have not thought it necessary to give the argument of the Republic 
in detail, as there are so many excellent accounts of it in existence 


done. It is not maintained, of course, that Plato is ever 
a mere reporter. He is clearly a dramatic artist, and 
arranges his material artistically. But he knew Sokrates 
well, and he wrote for people who knew Sokrates well, 
and the dialogues made use of in this sketch were probably 
all written before he came forward as a teacher of philo- 
sophy himself. If Plato's Sokrates is not meant for the 
real Sokrates, I find it very hard to imagine what he can 
be meant for. 



The Condemnation 

137. In 399 B.C. Sokrates was brought to trial by 
Anytos, the democratic leader, Meletos, a cc youthful and 
unknown" tragic poet 3 "with lanky hair, a scanty beard, 
and a hooked nose," 1 and Lykon, an even more obscure 
rhetorician. The indictment stated that he was guilty of 
not worshipping (i/o/uW) 2 the gods the State worshipped 
but introducing other new divinities, and further that he 
was guilty of corrupting the young by teaching them 
accordingly. In the Apology Plato gives us what profess 
to be the speeches delivered by Sokrates at his trial. It is 
not to be supposed that even here he is a mere reporter. 
It was usual for speeches to be carefully revised and 
adapted for publication, and no doubt Plato meant to do 
for Sokrates what other accused persons either did for 
themselves or more often had done for them by a profes- 
sional speech-writer. On the other hand it seems incredible 
that he should have misrepresented the attitude of Sokrates 
before the court or the general line of his defence. It is 
perfectly true, no doubt, that the Apology is not a defence 

1 Euthyphro, 2 b. 

2 The least inadequate translation of vopifav in its legal sense is 
" worship." The word does not refer primarily to " religious opinions/' 
but to the observance of certain current " uses " (vo/zcu), though Plato 
makes Sokrates take advantage of the secondary sense "think" in order 
to confuse Meletos (dfol. z6c). 


at all, but that makes it all the more characteristic of the 
man. Sokrates treats the accusation with contempt, and 
even goes out of his way to import things into the case 
that were hardly of a nature to conciliate the judges. 
That does not prove the Apology to be pure fiction, as it 
has been supposed to do. 1 Far from it. 

138. The actual conduct of the prosecution was 
entrusted to Meletos, who bungled it, according to Plato. 
By a skilful cross-examination Sokrates got him to admit 
that he believed him to be an out-and-out atheist, which 
was of course inconsistent with the indictment. In any 
case Sokrates did not stoop to defend himself against 
either the one charge or the other, though he showed 
himself more sensitive to the accusation of corrupting the 
youth, and offered to allow the fathers and elder brothers 
of his associates to give evidence on the point. He was 
found guilty, however, in spite of the failure of Meletos to 
make anything of the principal count in the indictment, 
which he does not seem to have understood himself. 
The majority was a considerable one, though Sokrates 
says he had expected it to be larger. He knew therefore 
that there was something else against him besides the 
trumpery charge of introducing new divinities, which he 
did not for a moment treat seriously. 

The penalty proposed by the accusers was death, but 
there is no reason to suppose Anytos really wished it to 
be carried out. By a very ingenious provision of the 
Athenian law, it was ordained that in cases of a certain 
class the condemned man should be allowed to propose an 
alternative sentence. The idea was that an adequate 
punishment would probably be arrived at in this way ; 
for the judges were bound to choose between the two 
penalties proposed, and could not suggest another them- 
selves. It was, therefore, the interest of the condemned 
man to propose something the judges would be likely to 
accept. There can be no doubt that if Sokrates had 

1 See the Introduction to Schanz's edition of the Apology with German 


proposed exile or imprisonment till he had paid a reasonable 
fine, everyone would have been satisfied, but he refused to 
do anything of the sort. That, he said, would amount to 
an acknowledgment of his guilt. If he had really to pro- 
pose what he thought he deserved, he would assess the 
penalty at free quarters in the Prytaneion at the public 
expense, an honour sometimes voted to Olympic victors 
and public benefactors. Ultimately, however, he proposed 
a fine of one mina, an inconsiderable sum, which his 
friends induced him to raise to thirty, offering to become 
surety for the payment. Plato was one of these friends, 
and this is the only act of his he has seen fit to put on 
public record. 

139. The judges were apparently incensed by this 
way of treating the court ; for they condemned Sokrates 
to death by a larger majority than that by which they had 
found him guilty. He then delivered a short address to 
those judges who had voted for his acquittal. He said 
that, even if death were the end of all things, it was no 
more an evil than a dreamless sleep, and few waking days 
are better than a night of that. He also hinted pretty 
plainly that, in his own belief, the soul was immortal, and 
that a good man had nothing to fear in the next life. And 
so he bade his judges farewell. "It is now time to depart, 
for me to die and for you to live. Which of us is going 
to meet the better lot, none knows but God." l 

The Alleged Offence. 

140. We have now to ask why Sokrates was charged 
with irreligion and why he was put to death. We must 
at once put aside the idea that it was for not believing the 

I It has actually been inferred from the Apology that "the historical 
Sukrates " had no fixed belief in immortality, and this has been used to 
discredit the Phaedo. I can only ask anyone who holds this view to 
read the passage aloud and see what effect it makes upon him. Of course 
Sokrates was addressing what was practically a public meeting, and he 
knew that few of his hearers held such beliefs, so there is some necessary 
reserve, but that is all. 


stories told about the gods. It is not likely that any 
educated man believed these, and uneducated people 
probably knew very little about them. 1 There was no 
church and no priesthood, and therefore the conception 
of religious orthodoxy did not exist. So far as mythology- 
was concerned, you might take any liberty. No one appears 
to have found fault with Aischylos for his Prometheus^ 
though, judged by modern standards, it is flat blasphemy. 
He did get into trouble for inadvertently revealing 
some Eleusinian formula, and the contrast is instructive. 
If it had been required of anyone that he should treat the 
stories about the gods respectfully, Aristophanes would 
not have survived Sokrates. He does not scruple to 
make fun of Zeus himself, and he represents Dionysos as 
a vulgar poltroon in a comedy which was actually part of 
the service of that very god and was presided over by his 
priest. In the Phaedrus (229 e sqq.} Sokrates is described 
as totally indifferent to the truth or falsehood of mythology, 
though he has the good taste to prefer the stories in their 
traditional form to the versions produced by the " homely 
wit" of rationalist historians. One thing he does indeed 
feel strongly, namely, that it is dangerous to repeat stories 
that ascribe untruthfulness and wickedness and strife to 
the gods, and in the Euthyphro (6 a) he does suggest that 
it is possibly for this that he is regarded as an innovator 
in religion. The suggestion is certainly not serious, how- 
ever, and even Euthyphro is not shocked, though he himself 
believes these stories and others stranger still. The truth 
is that belief in narratives of any kind formed no part of 
ancient religion; anyone might reject or accept such things 
as he pleased. Mythology was looked on as a creation of 
the poets, and c< poets tell many falsehoods." No one 
could be prosecuted for what we call religious opinions. 2 

141. Nor is it credible that the divine "voice" should 
have had anything to do with the prosecution. It is true 
that Euthyphro is represented as jumping at once to the 
conclusion that it had ; for that is the sort of thing he 

1 Arist. Poet. 1451 b, 25. * Cf. p. 76, n. 2. 


himself is interested in. At the same time, he makes it 
quite clear that, in his opinion, Sokrates need have no fear 
of a charge like that, though he must expect to be laughed 
at. 1 In the Apology Plato makes Sokrates himself say that 
the divine voice is presumably what Meletos has carica- 
tured and made the ground of the charge in his indictment, 
but the way he says it makes it quite clear that Meletos 
meant nothing of the sort and had said nothing about the 
"voice." 2 The Athenians might and did think Sokrates 
eccentric because of his voice and his trances, and, as 
Euthyphro says, such things are "easily misrepresented " 3 
and are apt to make people jealous. But the belief in 
"possession" (/caro/cor^) was much too firmly established, 
and cases of it were much too familiar, to allow of a charge 
of irreligion being based on anything of the kind. 4 The 
accepted view was that such things were a sort of disease 
which could be treated by " purifications," but even mad- 
ness and epilepsy were supposed to make the sufferer 
te holy " (fc/)oV). From the point of view of the ordinary 
Athenian, the irreligion would be on the side of anyone who 
treated the " voice " disrespectfully. 

142. It must also be remembered that the charge of 
introducing new divinities was no novelty; for it had been 
definitely formulated by Aristophanes a generation earlier. 
In the Clouds Sokrates announces that Zeus has been de- 
throned and Vortex reigned in his stead. He offers prayer 
to the Clouds and swears by Respiration, Chaos, and Air. 
It will be remembered that Diogenes of Apollonia held 
Air to be a "god." That being so, it is surely very signi- 
ficant that Aristophanes does not make the most distant 

Oi 3 b J f 

2 Jpology, 3 1 d. Professor Taylor's interpretation of the words o 8rj 
KOI . . . Iv ry y/oa<$ . . . eypai/'aro (Varia Socratica, i. p. 14) seems to 
me the only sound one. Sokrates says he supposes (&)) that Meletos 
meant the divine voice when he spoke of Scupovta, in the indictment. It is 
clear, then, that Meletos said nothing about it in Ms speech. 

8 The word i38ta/3oAa means no more. 

* The " voice " would no doubt strike the average Sao-tSat/wov as an 
ordinary case of yyacrr/>i/a>0ta. 


allusion to the " voice," though he must have known all 
about it, and it would lend itself admirably to comic treat- 
ment. The omission is the more striking, as there is an 
allusion to the trances of Sokrates (150). Xenophon is 
even more instructive. He says he got his information 
about the trial from Hermogenes, and we may be sure the 
religious Xenophon would be anxious to discover all he 
could about the meaning of this charge. He does not 
appear, however, to have got any definite explanation of 
it ; for he only gives it as his personal opinion that it must 
have been the "voice" on which the accusers chiefly relied, 
and it seems most probable that he is only repeating this 
from Plato's Apology and Euthyphro. At any rate, in his 
own Apology ^ he makes Sokrates speak about the " voice " 
very much as Plato does, and he makes him say, just like 
Euthyphro, that the Athenians are jealous of it as an ex- 
ceptional divine favour. In fact, everyone speculates about 
the meaning of the charge, and the one fact that stands out 
clearly is that no one not even the prosecutor seems to 
know it. It surely follows that the charge of introducing 
new divinities, though stated in the indictment, was neither 
explained nor justified at the trial. Such things were pos- 
sible in an Athenian dikastery, which was more like a public 
meeting than a court of justice. There was no judge to 
rule the prosecution irrelevant to the indictment. 

The Real Offence. 

143. But, if that is the true account of the matter, if 
follows further that this accusation was a mere pretext. 
That would explain why Meletos falls so easily into the 
trap laid for him by Sokrates, and substitutes the charge 
of atheism for that of introducing strange divinities. It 
will also make the conduct of the judges more intelligible. 
We know that a number of them, after voting for the 
acquittal of Sokrates on the charge brought against him, 
turned round and voted for the death sentence. That is 
partly to be explained, no doubt, by the attitude Sokrates 


took up in his second speech, but this will not explain it 
altogether. Death is surely an extreme penalty for con- 
tempt of court, and those judges must have believed 
Sokrates to be guilty of something. Everything becomes 
clear if we suppose that the real ground of the accusation 
could not for some reason be stated in the indictment, and 
that some of the judges thought it unfair to condemn a 
man for an offence with which he was not formally charged, 
even though they might believe him guilty of it. The 
defiant attitude of Sokrates would account for their change 
of mind in that case. 

Now we know that Sokrates had refused to obey the 
illegal orders of the Thirty, but we also know that he did 
not leave Athens. He was therefore suspect of incivisme^ 
but the amnesty made it impossible to charge him with a 
strictly political offence. Plato indicates in the clearest 

'-possible manner that Sdkrates really owed his death to his 
political attitude. There are two passages in which he is 
represented as criticising the democratic leaders of the fifth 
century, including Perikles, in a very severe manner. One 
of these is in the Gorgias, and there Kallikles,who is a demo- 

ucratic statesman, bluntly tells him (521 c) that, if he refuses 
to flatter the democracy instead of trying to make them 
see the error of their ways, he is in danger of being dragged 
into court by some sorry wretch, and then anything may 
happen to him. The other passage is in the Meno, where 

: Anytos himself is brought on the stage to give a similar 
warning. That is surely meant to be significant Anytos 
is not the chief interlocutor, and is apparently introduced 
solely for this purpose. After listening impatiently to the 
criticisms of Sokrates on the heroes of the democracy, he 
says (94 e), " I think, Sokrates, you are rather ready to 
abuse people, I should advise you, if there was any chance 
of your taking my advice, to be careful. Even in other 
cities, I fancy it is easier to do people a mischief than a 
good turn, and most decidedly it is so in this one." These 

*are very broad hints, and Plato set them down deliberately 
some time after the event. They can only mean that the 


real offence of Sokrates was his criticism of the democracy 
and its leaders. No one in Plato ever gives him a hint 
that he had better be careful not to talk about unauthorised 
divinities, as he frequently does, and still less does anyone 
suggest that the " voice" is a thing he would be wise in 
keeping to himself. 

144. From this point of view one of the most im- 
portant things in the Apology is the statement of Sokrates 
(39 d) that his countrymen will not be able to rid them- 
selves of criticism even if they put him to death. There 
are many who will take up the task of exposing them, and 
they will be more merciless inasmuch as they are younger. 
That is, to all intents and purposes, a plea of guilty to 
what the hints of Kallikles and Anytos suggest was the 
real ground of the accusation, namely, that Sokrates had 
fostered in young men that antidemocratic spirit which 
had led to the oligarchical revolutions. About half a 
century later Aischines put the matter quite bluntly. 
He says (i. 173) that the Athenians "put the Sophist 
Sokrates to death because he was believed to have 
educated Kritias," and less than ten years after his 
trial the Sophist Polykrates charged him, as we saw, 
with having educated Alkibiades. In fact, it looks as if 
Polykrates simply wrote the speech Anytos would have 
delivered at the trial, if the amnesty had not stood in the 
way. That the point was actually made by Meletos, a 
less responsible person, is strongly suggested by the 
allusion Sokrates makes in the Apology (33 a) "to those 
they say are my disciples.'' Xenophon also in the 
Memorabilia (i. 2, 12 sqq.} makes a point of saying 
that Kritias and Alkibiades were not really disciples of 

145. It is only fair to say that, from his own point 
of view, Anytos was not altogether wrong. Xenophon, 
indeed, attributes merely personal motives to him. He 
says in his Apology (29) that he was angry with Sokrates 
for telling him he ought to give his son a liberal educa- 
tion instead of bringing him up to his own business as a 


tanner. It is impossible to say what truth there may be 
in that, but in any case there were other reasons why 
Anytos should desire to remove Sokrates from Athens. 
He had undoubtedly been an uncompromising opponent 
jof the Periklean democracy, the radical vice of which, 
according to him, was that it denied the need for expert 
knowledge in politics. It would take the advice of experts 
on questions of shipbuilding or fortification ; but when a 
vital point of right or wrong in national policy had to 
be decided, anyone who chose to get up and speak was 
supposed to be as good a judge as anyone else. According 
J;o Plato, he went so far as to deny the title of statesman 
to the democratic leaders of his time, including Perikles. 
In the Republic we have an account of the democratic 
State, which is certainly meant to be a description of 
Athens in the fifth century, not of the humdrum bourgeois 
democracy of Plato's own time, and the description is 
by no means flattering. Of course the young men who 
followed Sokrates about would be far less impressed by 
his positive teaching than by this destructive criticism of 
existing institutions. They would be prejudiced against 
democracy to start with, and they would relish his attacks 
on it keenly. It is a fact that many of them became 
vulgar oligarchs and not statesmen. That is the tragedy 
of the situation. Sokrates was not responsible for it, but 
it existed all the same. Now Anytos and his friends were 
busily engaged in organising the restored democracy, and 
they could not afford to leave their work at the mercy of 
reaction. They had every reason to believe that the 
teaching of Sokrates was of a kind to imperil the con- 
stitution, and it is not surprising that they took steps 
accordingly. It must be remembered that they had pro- 
bably no desire to see Sokrates put to death, but it was 
natural they should wish to drive him into exile. In 
those circumstances we can easily understand why some 
of the friends of Sokrates thought it prudent to leave 
Athens for a time after his death. Even Plato went, 
though, as we shall see, he had held aloof from the 


oligarchical revolution in which his kinsmen were implicated, 
and though he had intended to enter public life under 
the restored democracy. Fortunately he found something 
better to do. 

The Pretext. 

146. Even assuming, however, that the charge of 
irreligion was a mere pretext, it must have been a colour- 
able one ; for the accusers ran the risk of being heavily 
fined if they did not secure a fifth of the votes. We must 
ask, then, whether there was anything that might be made 
to appear a justification of the charge, and on which a 
statesman like Anytos might rely to produce the right 
kind of prejudice against Sokrates. If we ask that ques- 
tion, we come at once upon the fact that in the very same 
year as Sokrates was tried Andokides appeared once 
more before the judges to explain his * connexion with 

^.the mutilation of the images of Hermes and the profana- 
tions of the mysteries sixteen years before. We find also 
that Anytos spoke in his favour, no doubt because his 

-revelations had been of service to the democratic party. 
We shall never know the truth about this old scandal, but 
the speech of Andokides is a precious document for the 
state of public feeling about it, not only at the time, but 
under the restored democracy. It is certain that, for 
the ordinary Athenian, the-- mutilation of the images was 
closely bound up with the profanation of the mysteries, 
and that both were supposed to be somehow/ directed 
towards the overthrow of the democracy. No doubt this 
was a mistake. The mutilation had probably nothing to 
do with the profanations of the mysteries, and the latter 
were obviously distorted in the popular imagination. It 
does-hot seem credible that some of the most gifted and 
enlightened men in Athens should have found it amusing to 

jDarody Eleusinian ritual, not once only or in a single place, 
though even that would be silly enough, but systemati- 
cally and in a number of private houses* On the other 
hand, the evidence that certain proceedings took place 


which were capable of being represented in that light is far 
too strong to be rejected, and conveys to a modern reader 
the idea that there may have been something resembling 
meetings of masonic lodges, exaggerated by public rumour 
into blasphemous mummeries of the most sacred rites. 

Now many of the judges must have known quite well 
that some of the most intimate associates of Sokrates 
were implicated in this business. There is no doubt, for 
instance, about Axiochos of Skambonidai, the uncle of 
Alkibiades and of Adeimantos son of Leukolophides, 1 
All three were denounced by Agariste, the wife of 
Alkmeonides, a high-born dame who had been the wife 
of one Damon before she married her kinsman. 2 This 
may very well be the same Damon whom Sokrates 
refers to as an authority on music. If that is correct, it is 
interesting to notice that one of the accused was called 
Taureas, and that is the name of the master of the 
palaistra in which Kritias introduced Charmides to 
Sokrates. 8 Further, if we remember that the banquet 
described in the Symposium is supposed to take place the 
very year the scandals occurred, it is suspicious that we 
find the names of Akoumenos, Eryximachos, and Phaidros 
among the persons inculpated/ Akoumenos was a cele- 
brated physician, and he has an unusual name. We do 
not know of anyone else who bore it. He was not 
present at the banquet, though his son Eryximachos, 
who was also a physician, is one of the speakers there. 
Phaidros is not an uncommon name, and we cannot be 
sure that Phaidros of Myrrhinous is meant. We are, 
however, told that he was an "associate" (era^o?) of 
Eryximachos, 6 and it is at the very least a remarkable 
coincidence that all three names should occur. In any 
case, we know that public interest in this old business had 

lr The record of the public sale of his confiscated goods still exists on 
inscriptions, where his name is given in full, 'Aftoxos 'AX/a/?ta8ov 
^Ka/jifitov&Tjs (Dittenberger, Syttoge 2 , 39, 41, 42, 45). 

2 Andok. i. 16. 8 Ib. i. 47 ; Plato, Charm. 153 a. 

*Andok. i. 15, 1 8, 35. 6 Plato, Phaedr. 268 a. 


just been revived, and that of itself would be sufficient to 
^create the atmosphere of prejudice required. Memories 
of the Clouds would do the rest. 

For reasons I have given, I do not think it likely that 
Sokrates was explicitly charged with this or any other 
particular offence against religion, but it was in everyone's 
mind, and there were circumstances enough in his life to 
connect him with it. It was certainly believed at Athens 
that he had taken part in religious rites of a strange kind ; 
for Aristophanes could count on his audience under- 
standing his allusions to them. Aischines wrote a dialogue 
in which Sokrates is represented as ^conversing with 
the Pythagorean Telauges. Plato represents him as 
full of Orphic ideas, though, as I have said 5 there is 
always a certain reservation which does not allow us to 
suppose he accepted them implicitly. I do not think it 
likely that his Pythagorean friends had much to do with 
this ; for, to all appearance, they had ceased to cc practise," 
and they had dropped the Orphic theory of the soul, 
which was just the thing that appealed most to Sokrates^ 
In fact, it is Sokrates who is represented as trying to* 
bring them back to an earlier form of Pythagorean 
belief. All this can hardly be fictitious. What motive 
could Plato have had for inventing it ? By his time 
Orphicism had hopelessly degenerated, so far as we can 
see, and it is not probable that it ever attracted him. 
Jn the youth of Sokrates things may well have been 
different. We know that the doctrine had been able to 
inspire a Pindar about the time Sokrates was born. 

The Death of Sokrates. 

147. Sokrates was not put to death at once. It was 
the festival of the Delian Apollo, and the ship the 

1 It will be seen where I am obliged to differ from my colleague 
Professor Taylor's conclusions in Varla Socratica, and I need not insist 
further on that. My agreement with him on other points will also be 


Athenians sent to Delos every year had just been solemnly 
garlanded the day before the trial. Now it was the law 
that the city should be kept free from the pollution of 
death at the hands of the public authority till the ship had 
gone to Delos and returned, and that sometimes took a 
M long time. So Sokrates had to spend a month in prison 
before his sentence could be carried out, and he passed 
that time in discussions with his friends, some of whom 
came from other parts of Hellas to bid him farewell. It 
would have been quite easy for him to escape at any time 
during this month, and his friends were ready to bear any 
expense that might be needful. But, as we have seen, 
Sokrates was a firm supporter of law, and he would not 
stoop to the inconsistency of making an exception in his 
own case. However unjust the sentence might be, it had 
been legally pronounced, and a good citizen could only 
submit. He owed everything to the laws of his country, 
and it was not for him to call them in question. 

In the Phaedo Plato has given an account of the last 
hours of Sokrates on earth. It would be difficult to match 
this narrative in the whole range of European literature, 
and it cannot be paraphrased. The last words of it are : 
" Such, Echekrates, was the end of our associate (eraZjOOff), 
a man, as we should say, the best and also the wisest and 
most righteous of his time.** 



148. A quite independent attempt at reconstruction 
was made by Demokritos. Like his contemporary Sokrates 
he faced the difficulties about knowledge raised by his 
fellow-citizen Protagoras and others, and like him he paid 
great attention to the problem of conduct, which had also 
been forced to the front by the Sophists. Unlike Sokrates, 
however, he was a voluminous author, and we can still see 
from his fragments that he was one of the great writers of 
antiquity. For us, however, it is almost as if he had 
written nothing, and we really know less of him than we 
do of Sokrates. That is because he wrote at Abdera, and 
his works were never really well known at Athens, where 
they would have had a chance of being preserved, like 
those of Anaxagoras and others, in the library of the 
Academy. It is not clear that Plato knew anything about 
Demokritos ; for the few passages in the Timaeus and else- 
where in which he seems to be reproducing him are easily 
explained by the Pythagorean influences that affected them 
both. Aristotle, on the other hand, knows Demokritos 
well ; for he too was an Ionian from the North. 

It is certain, nevertheless, that the Demokritean corpus 
(which included the works of Leukippos and others as 
well as those of Demokritos) continued to exist ; for the 
school maintained itself at Abdera and Teos down to 
Hellenistic times. It was therefore possible for Thrasyllos 
in the reign of Tiberius to produce an edition of the works 
of Demokritos arranged in tetralogies just like his edition 


of Plato's dialogues. Even that did not suffice to preserve 
them. The Epicureans, who ought to have studied the 
man to whom they owed so much, were averse to study of 
any kind, and probably did not care to multiply copies of 
a writer whose works would have been a standing testimony 
to the lack of originality that marked their own system. 

149. We know extremely little about the life of 
Demokritos. He belonged like Protagoras to Abdera 
in Thrace, a city which hardly deserves its proverbial 
reputation for dulness, seeing it could produce two such 
men. 1 As to the date of his birth, we have only con- 
jecture to guide us. In one of his chief works he stated 
that it was written 730 years after the fall of Troy, but we 
do not know when he supposed that to have taken place. 
There were several eras in use at the time and later. He 
also said somewhere that he had been a young man in the 
old age of Anaxagoras, and from this it was inferred that 
he was born in 460 B.C. That seems rather too early, 
however ; for it is based on the assumption that he was 
forty years old when he met Anaxagoras, and the 
expression "young man" suggests something less than 
that. Further, we have to find room for Leukippos 
between him and Zeno. If Demokritos died, as we 
are told, at the age of ninety or a hundred, he was in 
any case still living when Plato founded the Academy. 
Even on purely chronological grounds, then, it is wrong 
to class Demokritos with the predecessors of Sokrates, 
and it obscures the fact that, like Sokrates, he tried to 
answer his distinguished fellow-citizen Protagoras. 2 

150. Demokritos was a disciple of Leukippos, and we 

1 It has been plausibly suggested that the reputation of the Abderites 
may have arisen from some satirical remark of Demokritos himself. 
The other side of the same thing may be represented by the view 
of Demokritos as "the laughing philosopher," which appears for the 
first time in Horace. 

2 As has been pointed out above (p. 112, #. 2), the stories which 
make Protagoras a disciple of Demokritos are based on the illusion 
that Protagoras was a contemporary of Plato. 


have contemporary evidence, that of Glaukos of Rhegion, 
that he also had Pythagoreans for his teachers. A later 
member of the school, Apollodoros of Kyzikos, says he 
learnt from Philolaos, and it seems quite likely. That 
accounts for his geometrical knowledge, and also, we shall 
see, for other features in his system. We know, too, 
that Demokritos spoke of the doctrines of Parmenides 
and Zeno in his works. These he would come to know 
through Leukippos. He mentioned Anaxagoras, as we 
have seen, and he appears to have said that his theory 
of the sun and moon was not original. That may refer to 
the explanation of eclipses, which was generally attributed 
at Athens, and no doubt in Ionia, to Anaxagoras, though 
Demokritos would, of course, know it to be Pythagorean. 

He is said to have visited Egypt, but there is some 
reason for believing that the fragment in which this is 
mentioned (fr. 298 b) is a forgery. There is another (fr. 
1 1 6) in which he says : "I went to Athens and no one 
knew me." If he said that, he meant no doubt that he 
had failed to make such an impression as his more brilliant 
fellow-citizen Protagoras had done. On the other hand, 
Demetrios of Phaleron said Demokritos never visited 
Athens at all, so this fragment may be a forgery too. In 
any case, most of his time must have been spent in study, 
teaching and writing at Abdera. He was not a wandering 
Sophist of the modern type, but the head of a regular 

The real greatness of Demokritos does not lie in the 
theory of atoms and the void, which he seems to have 
expounded much as he had received it from Leukippos. 
Still less does it lie in his cosmological system, which is 
mainly derived from Anaxagoras. He belongs to another 
generation altogether than these men, and he is not 
specially concerned in finding an answer to Parmenides. 
The question he had to deal with was that of his own 
day. The possibility of science had been denied and the 
whole problem of knowledge raised by Protagoras, and 
that had to be met Further, the problem of conduct 


had become a pressing one. The originality of Demokritos 
lay, then, precisely in the same directions as that of 

Theory of Knowledge. 

151, Demokritos followed Leukippos in giving a 
purely mechanical account of sensation, and it is probable 
that he is the author of the detailed atomist doctrine 
on this subject. As the soul is composed of atoms like 
everything else, sensation must consist in the impact 
of atoms from without on the atoms of the soul, and the 
organs of sense must be simply " passages" (nopai) 
through which these atoms are introduced. It follows 
that the objects of vision are not strictly the things we 
suppose ourselves to see, but the cc images " ($e//ceXa, 
e?<Ja)Xa) that bodies are constantly shedding. The image in 
the pupil of the eye was regarded as the essential thing in 
vision. It is not, however, an exact likeness of the body 
from which it comes ; for it is subject to distortion by 
the intervening air. That is why we see things in a 
blurred and indistinct way at a distance, and why, if the 
distance is very great, we cannot see them at all. If there 
were no air, but only the void, between us and the objects 
of vision, this would not be so ; cc we could see an ant 
crawling on the sky." Differences of colour are due 
to the smoothness or roughness of the images to the 
touch. Hearing is explained in a similar way. Sound is 
a stream of atoms which flow from the sounding body and 
cause motion in the air between it and the ear. They 
therefore reach the ear along with those portions of the air 
that resemble them. The differences of taste are due 
to differences in the figures (ei'&y, o^^ara) of the atoms 
which come in contact with the organs of that sense, and 
smell was similarly explained, though not in the same 
detail. In the same way, touch, regarded as the sense 
by which we feel hot and cold, wet and dry, and the like, 
is affected according to the shape and size of the atoms 
impinging upon it. 


Aristotle says Demokritos reduced all the senses to 
that of touch, and that is quite true if we understand 
by touch the sense that perceives such qualities as shape, 
size and weight. This, however, must be carefully dis- 
tinguished from the special sense of touch which has just 
been described. To understand this point, we must go 
on to consider the doctrine of "trueborn" and "bastard" 

152. It is here that Demokritos comes sharply into 
conflict with Protagoras, who had declared all sensations 
to be equally true for the sentient subject. Demokritos, 
on the contrary, regards all the sensations of the special 
senses as false, inasmuch as they have no real counterpart 
outside the sentient subject. In this he is of course true 
to the Eleatic tradition on which the atomic theory rests. 
Parmenides had said expressly that taste, colours, sound, 
and the like were only " names " (ovojuLara), and it is quite 
likely Leukippos said something of the same sort, though 
there is no reason to believe he had elaborated a theory on 
the subject. Coming after Protagoras as he did, Demo- 
kritos was bound to be explicit on the point. His doctrine 
has fortunately been preserved to us in his own words. 
"By use (yoV&>)," he said (fr. 125), "there is sweet, by 
use there is bitter ; by use there is warm and by use there 
is cold ; by use there is colour. But in sooth (ere#) there 
are atoms and the void." In fact, our sensations represent 
nothing external, though they are caused by something 
outside us, the true nature of which cannot be apprehended 
by the special senses. That is why the same thing is 
sometimes felt as sweet and sometimes as bitter. <c By the 
senses," Demokritos said (fr. 9), " we in truth know 
nothing sure, but only something that changes according 
to the disposition of the body and of the things that enter 
into it or resist it." We cannot know reality in this way; 
for cc truth is in the depths" (fr. 117). It will be seen that 
this doctrine has much in common with the modern dis- 
tinction between the primary and secondary qualities of 


153. Demokritos, then, rejects sensation as a source of 
knowledge just as the Pythagoreans and Sokrates did ; but, 
like them, he saves the possibility of science by affirming 
that there is a source of knowledge other than the special 
senses. "There are/' he says (fr. n), "two forms of 
knowledge (ywo/w?), the trueborn (yw/o-ii/) and the bastard 
(o-Korirf). To the bastard belong all these ; sight, hearing, 
smell, taste, touch. The trueborn is quite apart from 
these." That is the answer of Demokritos to Protagoras. 
He had said that honey, for instance, was both bitter and 
sweet, sweet to me and bitter to you. In reality it was 
"no more such than such" (ovSev ju.$X\ov TOIOV *; TO?OI/). 
Sextus Empiricus and Plutarch tell us expressly that Demo- 
kritos argued against Protagoras, and the fact is therefore 
beyond question. 

At the same time, it must not be overlooked that Demo- 
kritos gave a purely mechanical explanation of this true- 
born knowledge just as he had done of the bastard. He 
held, in fact, that the atoms outside us could affect the atoms 
of our soul directly without the intervention of the organs of 
sense. The atoms of the soul were not confined to any 
particular parts of the body, but permeated it in every 
direction, and there was nothing to prevent them from 
having immediate contact with the external atoms, and so 
coming to know them as they really are. The " true-born 
knowledge" is, after all, of the same nature as the 
" bastard," and Demokritos refused, like Sokrates, to 
make an absolute separation between sense and thought. 
"Poor Mind," he makes the senses say (fr. 125), " it is 
from us thou hast got the proofs to throw us with. Thy 
throw is a fall." 1 The " true-born " knowledge is, after 
all, not thought, but a sort of inner sense, and its objects 
are like the "common sensibles" of Aristotle. 

1 54. As might be expected from a follower of the 

Pythagoreans and Zeno, Demokritos busied himself with 

the problem of continuity. In one remarkable passage 

(fr. 155) he states it in this form : " If a cone is cut by a 

1 Cp. p. 113, . 2. 


plane parallel to its base, what are we to think of the 
surfaces of the two sections ? Are they equal or unequal ? 
If they are unequal, they will make the cone uneven ; for 
it will have many step-like incisions and roughnesses. If 
they are equal, then the sections will be equal, and the 
cone will have the properties of a cylinder, which is com- 
posed of equal, not unequal, circles. Which is most 
absurd." From a remark of Archimedes x it appears that 
Demokritos went on to say that the volume of the cone 
was a third of that of the cylinder on the same base and 
of the same height, a proposition first demonstrated by 
Eudoxos. It is clear, then, that he was engaged on 
problems such as those which ultimately gave rise to the 
infinitesimal method of Archimedes himself. Once more 
we see how important the work of Zeno was as an intel- 
lectual ferment. 

theory of Conduct* 

155. The views of Demokritos on conduct would be 
even more interesting than his theory of knowledge if we 
could recover them completely. It is very difficult, how- 
ever, to be sure which of the moral precepts attributed to 
him are genuine. There is no doubt that the treatise on 
Cheerfulness (TLepl evOvjuiiw') was his. It was freely used by 
Seneca and Plutarch, and some important fragments of it 
have survived. 

It started (fr. 4) from the principle that pleasure and 
pain (Wp\J/t? and arep-^in) are what determine happiness. 
This means primarily that happiness is not to be sought 
for in external goods, " Happiness dwelleth not in herds 
nor in gold ; the soul is the dwelling-place of the daimon" 
(fr. 171). To understand this, we must remember that the 
word Saifjiciw, which properly meant a man's guardian spirit, 
had come to be used almost as the equivalent of " fortune.'' 
It is, as has been said, the individual aspect of rv^rjy and 
the Greek word we translate by " happiness" (evSawovla) 
is based on this usage. On one side of it, then, the 

*Cf. Diels, Fors* ii. p. 90 n. 


doctrine of happiness taught by Demokritos is closely 
related to that of Sokrates, though it lays more stress on 
pleasure and pain. " The best thing for a man is to pass 
his life so as to have as much joy and as little trouble as 
may be" (fr. 189). 

This is not, however, vulgar hedonism. The pleasures 
of sense are just as little true pleasures as sensations are 
true knowledge. " The good and the true are the same 
for all men, but the pleasant is different for different people" 
(fr. 69). Further, the pleasures of sense are of too short 
duration to fill a life, and they easily turn into their 
opposite. We can only be sure of having an excess of 
pleasure over pain if we do not seek our pleasure in what 
is " mortal" (fr. 189). 

What we have to strive after is "well-being" (eueo-rco) 
or "cheerfulness" (euQvfd^ and that is a state of the soul. 
To attain it, we must be capable of weighing, judging, and 
distinguishing the value of different pleasures. Just like 
Sokrates, Demokritos laid down that " ignorance of the 
better" (fr. 83) was the cause of failure. Men put the 
blame on fortune, but that is only an "image" they have 
invented to excuse their own ignorance (fr. 119). The 
great principle which should guide us is that of "symmetry" 
or " harmony." That is, no doubt, Pythagorean. If we 
apply this test to pleasures, we may attain to " calm," calm 
of body, which is health, and calm of soul, which is cheer- 
fulness. That is to be found chiefly in the goods of the 
soul. " He who chooses the goods of the soul chooses 
the more divine ; he who chooses the goods of the c taber- 
nacle' (i.e. the body) 1 chooses the human" (fr. 37). 

156. For our present purpose it is not necessary to 
discuss the cosmology of Demokritos in detail. It is 
thoroughly retrograde and proves, if proof were needed, 
that his real interests lay in another direction. He had 

1 This use of ovc^vos for the body (found also in S. Paul, 2 Cor. v. i) 
is probably Pythagorean, and connected with the representation of human 
life as a Travrjyvpis or " fair." Our bodies are our temporary " booths." 


inherited the theory of atoms and the void from Leu- 
kippos, who was the real man of genius in this field, and 
he was content for the rest to adopt the crude Ionic 
cosmology as Leukippos had done. Yet he must have 
known the more scientific system of Philolaos. The 
knowledge of the earth's spherical shape was widely spread 
by the days of Demokritos, and Sokrates is represented 
in the Phaedo (108 e) as taking it for granted. For 
Demokritos the earth was still a disc. He also followed 
Anaxagoras in holding that the earth was supported on 
the air " like the lid of a trough," another view which 
Sokrates rejects with emphasis. On the other hand, 
Demokritos appears to have made valuable contributions 
to natural science. Unfortunately our information is far 
too scanty to permit even an approximate reconstruction 
of his system. The loss of the complete edition of his 
works by Thrasyllos is perhaps the most deplorable of 
our many losses of this kind. It is probable that they 
were left to perish because Demokritos came to share in 
the discredit that attached itself to the Epicureans. What 
we have of him has been preserved mainly because he was 
a great coiner of telling phrases, and these have found 
their way into anthologies. That is not the sort of 
material we require for the interpretation of a philosophic 
system, and it is very doubtful whether we know some of 
his deepest thoughts at all. At the same time, we cannot 
help feeling that it is mainly for their literary merit that 
we regret the loss of his works. He seems to stand apart 
from the main current of Greek philosophy, and it is to 
that we must now return. From our point of view, the 
only important fact about Demokritos is that he, too, saw 
the need of an answer to Protagoras. 

BOOK /// 



Plato s Early Life 

157. If the Epistles are genuine and some of the 
greatest scholars and historians hold they are we know 
more of the life of Plato than of any other ancient philo- 
sopher. 1 Even apart from the Epistles, we know a good 
deal. Besides what we may infer from the dialogues, we 
have one or two statements resting on the authority of 
Hermodoros, who was a member of the Academy in 
Plato's time, and these give us certain fixed points to start 
from. The later Lives are almost entirely mythical. It is 
conceivable that they may contain one or two stray facts 
derived from older sources now lost, but their general 
character is such that it is safer to neglect them in the 
first instance. The Epistles^ on the other hand, are free 
from this mythology, which is the more remarkable as 
Plato's own nephew, Speusippos, already credited him with 
a miraculous birth. If, then, the Epistles are forgeries, 
they are at least the work of a sober and well-informed 

1 The genuineness of the Epistles has been maintained by scholars like 
Bentley and Cobet, and by historians like Grote and E. Meyer. In 
practice most accounts of Plato really depend on them, though that is 
disguised by the custom of referring instead to Plutarch's Life of Dion. 
Plutarch, however, is obviously dependent on the Epistles for most, if not 
all, of what he tells us ; so this is an illegitimate evasion. I should ^add 
that the First Epistle stands by itself. In my judgement, it has got into 
its present place by mistake. It is a genuine fourth-century letter, but 
I do not think the writer, whoever he was, meant to pass for Plato at all, 
I do not think either that he was Dion or meant to pass for Dion. 


writer, whose use of the Attic dialect proves him to have 
been Plato's contemporary. It would have been impos- 
sible to find anyone fifty years later who could handle the 
language as he does. 1 Even the oldest and most successful 
of the spurious dialogues betray themselves at every turn. 
We may, indeed, go so far as to say that the supposed 
forger of the Epistles must have been a man of almost 
unparalleled literary skill, or he could not have reproduced 
so many of the little peculiarities that marked Plato's style 
at the very time of his life to which the Epistles profess to 
belong, though with just those shades of difference we 
should expect to find in letters as contrasted with more 
elaborate literary work. I believe that all the letters of any 
importance are Plato's, and I shall therefore make use of 
them. As, however, there are still eminent scholars who 
are not convinced, I shall warn the reader when I have 
occasion to do so. 

158. Plato was born in 428/7 B.C., more than a year 
after Perikles died and just before Gorgias came to Athens 
for the first time. We learn from a poem quoted in the 
Republic (368 a) and addressed to his brothers, Adeimantos 
and Glaukon, that his father, Ariston, was a man of dis- 
tinction. He must have died when Plato was a child ; 
for his wife, Periktione, afterwards married Pyrilampes, 
whose son by her, Antiphon, was in his youth an associate 
of Pythodoros son of Isolochos, who had been a disciple 
of Zeno. Adeimantos and Glaukon must have been older 
than Plato. The idea that they were younger is based on 
a misunderstanding of the Republic. It is assumed that 
Plato could not talk as he does there except to younger 
brothers, and it is forgotten, as usual, that Sokrates, not 
Plato, is the speaker. In the Apology (34 a) Sokrates says 
Adeimantos should have been called to give evidence 
whether Plato had got any harm from associating with 
him, and this implies that Adeimantos was so much older 
as to stand in loco parentis to his brother. Further, we 

1 After the rise of Atticism It might have been just possible, but we 
know the Epistles existed before that. 


learn from the poem quoted in the Republic that both 
Glaukon and Adeimantos had won distinction in the battle 
of Megara. It is natural, in the absence of further qualifi- 
cations, to suppose that the battle of 424 B.C. is meant, 
though we cannot be quite certain. In any case, if both 
the brothers won distinction in the same battle, they cannot 
have differed widely in age. It may be added that it would 
not have been in accordance with Plato's usual practice to 
introduce his brothers in the Republic if they had been still 
living when that dialogue was written. Xenophon (Mem. 
iil 6, i) tells a story of how Glaukon was restrained by 
Sokrates from speaking in the Assembly before he had 
reached the legal age of twenty. Sokrates did that by 
asking him a series of questions about Athenian finance 
and the national defences, and it is impossible to read these 
questions without feeling that Xenophon conceived the 
incident to have taken place some time before the occupa- 
tion of Dekeleia in 41 3 B.C. It is true that he says Sokrates 
was interested in Glaukon because of Charmides and Plato, 
but that may be a slip. Charmides was at least twenty 
years older than Plato, who would, perhaps, be too young 
to attract the attention of Sokrates much before 413 B.C. 
The slip, however, if it is one, is explicable enough in a 
writer so careless of chronology as Xenophon, and cannot 
outweigh the other presumptions. As to Charmides, we 
know that Sokrates made his acquaintance four or five 
years before Plato was born, so the mention of his name is 
quite appropriate. 

The family of Plato's mother, Periktione, was also highly 
distinguished, and traced its descent to Dropides, the friend 
and kinsman of Solon. She herself was the cousin of 
Kritias and the sister of Charmides, son of Glaukon, and 
the fact that Glaukon bore the name of his maternal grand- 
father affords a further presumption that he was the second 
son. As we are told in the Charmides (158 a) that Pyri- 
lampes was the maternal uncle of Charmides, we jmust 
assume that Periktione was his niece, and that he married 
her when she was left a widow by the death of Ariston. 


That would be in accordance with Athenian usage. The 
last we hear of Pyrilampes is that he was wounded in the 
battle of Delion, but Periktione reached a great age ; for 
it appears from Epistle xiii. (361 e) that she was still 
living in 366/5, though her death was expected. 1 The 
importance of all this is that it enables us to identify the 
Glaukon and Adeimantos of the Parmenides with those of 
the Republic, and also to fix the supposed date of the 
latter dialogue before the departure of Polemarchos for 
Thourioi instead of after his return. That explains how 
Kephalos is still alive, and how Lysias, though present, 
does not take any part in the conversation. We shall see 
that a good deal depends on this. 

Plato was undoubtedly proud of his illustrious kinsmen, 
and he introduces them over and over again in his writings. 
The 'opening scene of the Charmides is a glorification of 
the whole connexion. It recalls the praises bestowed on 
the house of Dropides by Solon and Anakreon, the 
youthful beauty and modesty of Charmides, and the fair 
stature of Pyrilampes, who was accounted the tallest and 
handsomest man in Asia when he went on an embassy 
to the King. The elder Kritias plays an important 
part in the Timaeus and in the dialogue called by his 
name. 2 Plato's reticence about himself stands in strik- 
ing contrast to the way he celebrates the older members 
of his family, especially as their memory was by no 
means popular at the time he wrote. I have called 
attention elsewhere 3 to the dramatic skill with which he 
keeps the shadow of the Revolutions from falling on 
his picture. His dialogues are not only a memorial to 
Sokrates, but also to the happier days of his own family. 
Plato must have felt the events of the end of the fifth 

1 This has been used as an argument against the genuineness of 
Epistle xiii., but it involves no impossibility, even if Adeimantos and 
Glaukon fought at Megara in 424 B.C. Athenian girls married very 
young, and it was a long-lived family. See the genealogical table in the 

2 See p. 3 38, #. i . 8 See my edition of the Phaedo, Introduction, IX 


century keenly^ but he is so careful to avoid anachronisms 
in these dialogues that no one could ever guess from them 
that they were written after Kritias and Charmides had 
met with a dishonoured end. 

159. The statement that Plato only made the ac- 
quaintance of Sokrates when he was twenty does not rest 
on the authority of Hermodoros, and is quite incredible. 
The nephew of Charmides must have known Sokrates 
ever since he could remember. It does not follow, how- 
ever, that he was one of the inner circle of disciples, and it 
is not very likely. It seems rather to have been the death 
of Sokrates that converted him to philosophy. That, at 
any rate, is the impression left by Epistle vii. There we 
are told quite distinctly (324 b) that he had looked 
forward to a political career. Kritias and Charmides for 
they are no doubt meant suggested that he should enter 
public life under the Thirty, but he was disgusted by their 
excesses, which made the former constitution seem like 
gold by comparison (324 d). In particular, he was shocked 
by the treatment of Sokrates in the affair of Leon of 
Salamis (111). When the democracy was restored, 
Plato thought once more of a political career, but the trial 
and death of Sokrates convinced him that this was im- 
possible in the Athens of his time. He could do nothing, 
he says (325 d), without joining a party, and neither of 
the existing parties could satisfy him. It was just as well. 
Athenian politics at this time were of no serious impor- 
tance, and, as he says in another letter (v. 322 a), "Plato 
was born late in the day for his country." He did, how- 
ever, find an opening in politics later, and on a much 
wider stage. 

1 60. It has become a commonplace to say that Plato's 
birth and connexions would incline him from the first to 
the oligarchic side, but nothing can be more untrue. The 
traditions of the family were rather what we should call 
"Whiggish," as is shown by the stress laid on its con- 
nexion with Solon. Even at the time of the brief domina- 
tion of the Four Hundred, Kritias was an opponent of the 

2io PLATO 

oligarchical extremists. Charmides became an oligarch at 
a later date, when he had been ruined by the war, but he 
did not at first take any part in politics. According to 
Xenophon it was Sokrates that urged him to overcome 
his natural shyness and enter public life (Mem. iii. 7). 
Moreover, Plato's stepfather and grand-uncle, Pyrilampes, 
was a friend of Perikles and a convinced democrat. It 
was not for nothing that he called his son Demos. It 
appears also from the Republic that Glaukon and Adei- 
mantos were intimate with the family of Kephalos, the 
wealthy stranger whom Perikles had persuaded to settle in 
Peiraieus. They were friends of his son Polemarchos, 
who afterwards met his death at the hands of the Thirty. 
In fact, so far as we can see, Plato's early upbringing 
would predispose him in favour of the Periklean regime. 
He says in the Seventh Epistle (325 b) that he was at first 
impressed by the moderation of the restored democracy, 
and such a thought would not be likely to occur to one 
brought up in the oligarchic camp. We can understand, 
then, why Plato's own judgment of democracy, as we have 
it in the Statesman and the Laws, is not nearly so harsh as 
that he puts into the mouth of Sokrates. 

161. Plato tells us in the Phaedo (59 b) that he was 
ill at the time Sokrates was put to death, and was therefore 
unable to be present. He had been in court at the trial, 
as we know from the Apology (38 b), and had offered with 
others to become surety for the payment of a fine, if the 
court would accept that penalty. After the death of 
Sokrates, Hermodoros said that he retired to Megara with 
some of the other Sokratics. We have seen ( 145) that 
they may well have been in some danger. Eukleides 
would of course receive them gladly, but we have no 
indication of the length of their stay with him. The later 
Lives attribute extensive travels to Plato, most of which 
are plainly apocryphal. It is probable, though by no 
means certain, that he visited Egypt. In the Laws 
(656 e) he speaks as if he had seen the monuments, 
he shows some knowledge of Egyptian methods of 


education (819 b). In any case, It was not to study 
mathematics he went there ; for we know that his opinion 
of Egyptian science (747 c) was by no means so favourable 
as that he expresses of Egyptian art. If he was in Egypt, 
it is likely that he also went to Kyrene to visit the mathe- 
matician Theodoros, who was a friend of Sokrates, but he 
may equally well have made his acquaintance at Athens, 
where he was teaching just before the death of Sokrates. 
All this, however, is extremely doubtful, and the earliest 
definite fact we know is that he visited Italy and Sicily for 
the first time when he was forty years old (Ep. vii. 324 a). 
It is likely that he wished to make the acquaintance of the 
distinguished Pythagoreans who were becoming powerful 
once more in these parts, and it was probably through 
them that he made the acquaintance of Dion, who was 
then about twenty. That brought him to the court of 
Dionysios L at Syracuse, where he was disgusted by the 
luxurious life he had to lead. The story goes that his 
freedom of speech offended Dionysios, who handed him 
over to the Spartan ambassador Pollis, who sold him as a 
slave at Aigina. His life was even in danger, but he was 
ransomed by a man of Kyrene named Annikeris. If this 
story is true, it is strange that it is not mentioned in the 
Seventh Epistle. Perhaps Plato may have thought it 
irrelevant in what is really a narrative of his relations with 
Dion and the younger Dionysios. A forger would hardly 
have omitted it, if the story had been current, but Plato 
himself might conceivably do so. In any case, he was 
back at Athens before long. 

162. At this time Plato was just over forty, and 
Sokrates had been dead twelve years. One good reason 
for holding he did not spend these years in continuous 
travel, as the later accounts suggest, is that he must have 
written a very considerable number of his dialogues already. 
Without deciding anything as to the order in which they 
were composed, we are able to say with some confidence 
that the Euthyphro, Apology, Crito y Charmides, Laches, Lysis, 
EuthydemuS) Protagoras, Gorgias, and Meno at least were all 


composed before Plato was forty. 1 That is about one 
dialogue a year, assuming that he wrote none of them 
before the death of Sokrates. If we remember that the 
great tragedians often brought out four plays in one 
year, that will not seem an excessive rate of production, 
and I have little doubt that the Symposium and Phaedo 
were also written by this date, and the Republic at least 
well advanced. In any case, it seems clear that all these 
works must have been completed before the foundation 
of the Academy, and I think we may take it that the 
Phaedrus is not very much later. In all these dialogues 
the dramatic interest seems to outweigh every other, 
except in some portions of the Republic. Plato's dramatic 
power, though often acknowledged in words, is seldom 
done justice to. He had a marvellous gift of assum- 
ing the most diverse personalities, and this gift is seen at 
its best in the Symposium^ which is certainly not one of 
the earliest dialogues, but goes with the Phaedo and the 
Republic. I cannot imagine that the man who could speak 
at will in the character of Protagoras or Gorgias, or Aristo- 
phanes or Alkibiades, without revealing anything of his 
own personality, should simultaneously, either voluntarily 
or involuntarily, have used Sokrates as a mask for himself. 
I do not therefore think it possible to learn much of Plato's 
own inmost thoughts from any of these dialogues, and I 
believe we have a perfectly serious statement to that effect 
in the Second Epistle. There he says (3140) : " There is 
no writing of Plato, nor will there ever be. What go by 
the name really belong to Sokrates turned young and 
handsome." The dialogues, in fact, profess to be pictures 
of a generation that had passed away, and that I believe 
them in the main to be. I do not think it likely that 
Plato had as yet anything that could rightly be called a 
philosophy of his own. He seems to have been one of 
those men whose purely intellectual development was late 

1 I have ventured to assume the results of the stylistic researches 
inaugurated by Lewis Campbell in 1867. It would take too long to 
discuss them here. 


and continued into old age. At first the artistic interest 
was paramount ; the purely philosophical does not gain 
the upper hand till his artistic gift declined. It is only in 
certain parts of the Republic and the Phaedrus that I can 
detect anything so far that seems to be Platonic rather than 
Sokratic, and I attribute that exception to the fact that 
Plato was about to open the Academy. The higher edu- 
cation of the Guardians seems to be a programme of the 
studies that were to be pursued there; and, as we shall 
see, Plato is not quite at his ease in making Sokrates speak 
of one of them, namely, solid geometry. Sokrates had 
proposed to take astronomy immediately after plane geo- 
metry, but he corrects himself and interpolates geometry 
of three dimensions, to which Glaukon objects that this 
has not yet been invented. It had been invented by 
Plato's time, and by a friend of his own. The awkward- 
ness he evidently feels in introducing it is to my mind 
very instructive. If he had already attributed to Sokrates 
all manner of scientific interests that were really foreign to 
him, why should he boggle at solid geometry ? 

Foundation of the Academy. 

163. The foundation of the Academy by Plato soon 
after his return to Athens was not only the most important 
event in his life, but also in the history of European science. 
The idea was no doubt suggested to him in the first place 
by the school of Eukleides at Megara, and by what he had 
seen of the Pythagorean societies in southern Italy. The 
name Academy is derived from a gymnasium outside the 
walls of Athens, which had been laid out as a public park 
by Kimon. Here Plato had a house and garden, and this 
remained for long the seat of the school, though it moved 
into the town after the siege of Athens by Sulla in 86 B.C., 
and continued to exist there till it was disestablished and 
disendowed by Justinian in 529 A.D. Like all societies of 
the kind, it was organised as a religious guild. It had its 
chapel, dedicated to the Muses, and its sacrifices at stated 


times. The members lived for the most part a common 

From the first the Academy attracted a large number of 
young men, many of whom became distinguished after- 
wards. It is to be observed that they came from almost 
every part of the Hellenic world. That is one of the 
things that distinguish the fourth century from the fifth. 
In the fifth century, the youth of Athens got their higher 
education from a number of distinguished foreigners who 
paid flying visits from time to time; in the fourth, the 
youth of all Hellas came to Athens to sit at the feet of 
two Athenian citizens, Isokrates and Plato. Athens had, 
in fact, become * the school of Hellas." It is of interest 
to note further that a goodly number of these youths came 
from the North, and especially from the Greek colonies in 
Thrace and on the Black Sea. That may have been due 
in some measure to the existence of a mathematical school 
at Kyzikos, of which Eudoxos was the head. At any rate, 
Eudoxos transferred himself and his school bodily to the 
Academy, which is all the more remarkable as he did not 
by any means see eye to eye with Plato on mathematical 
and astronomical subjects. It can hardly be an accident 
that Ionia proper is so poorly represented in the Academy, 
so far as we know who composed it. The lonians had 
rejected Pythagorean science, partly no doubt because it 
was mixed up with mysticism. The School of Demokritos 
continued to exist at Teos down to Hellenistic times. In 
Plato, Euthydemos and Dionysodoros come from Chios, 
and Euboulides, the adversary of Aristotle, was a Milesian. 
That is all we can say of Ionia till the time when Epicurus 
of Samos once more brought the old Ionic tradition to 
Athens, where it had been unrepresented since the days of 

It is of the utmost importance to remember that Plato's 
real teaching was given in the Academy, and that even his 
later dialogues only contain what he thought fit to give to 
a wider public in order to define his attitude to other 
schools of philosophy. This fact, which is often over- 


looked, accounts for a great deal of the difficulty we feel 
in passing from Plato to Aristotle. We seem to be in a 
different world altogether, and that is natural ; for we 
have neither Plato's lectures nor (except in fragments) the 
published works of Aristotle, and we are thus comparing 
two quite different things. If we only had Plato's lecture 
on The Good and the Protreptikos of Aristotle, we should 
get a very different impression. As it is, we may fairly 
assume that Plato's lectures had far more resemblance to 
Aristode's than to his own dialogues. 

164. It will help us considerably to understand the 
purpose of the Academy if we first consider what Plato 
meant by the word cc philosophy." In Ionia it had been 
used of a more or less scientific curiosity which led men 
to visit strange lands and note their usages. It may 
have been applied also to the researches (larroplrf) of the 
Milesians, but there is no evidence of that. It was in 
all probability Pythagoras that first gave it the deeper 
meaning of science " touched with emotion," and it was 
certainly in the Pythagorean community that it came 
to be regarded as a " way of life." For Sokrates too, 
according to Plato, philosophy had been above all things 
a life. At Athens, however, the word was current in 
a vaguer and shallower sense, derived probably from the 
Ionian usage. It had, in fact, a range of meaning some- 
thing like that of our word " culture." The great teacher 
of philosophy in this sense was Isokrates, the only 
Athenian of the time whose influence was at all com- 
parable to Plato's. Much that has been written about 
the attitude of these two men to one another is extremely 
fanciful, but the main facts are clear enough. It will be 
well to state them briefly here, for it is really necessary to 
understand Isokrates if we are to estimate Plato aright. 

Plato and Isokrates. 

165. One thing was common to both men, and that 
was an intense belief that the only remedy for the ills 


of Hellas was enlightenment, though they differed enor- 
mously as to the kind of enlightenment required. There 
is a striking passage at the end of the Phaedrus^ where 
Sokrates is made to contrast Isokrates with mere professional 
advocates like Lysias, He says : 

Isokrates is still young, but I am ready to tell you what I 
presage for him. ... I think that, so far as natural gifts go, 
he is capable of higher things than the speeches of Lysias, 
and that his character is more nobly tempered. It would be 
no wonder, then, as he grows older, if, even in composing 
speeches, which is the task he is now engaged on, he 
should make all who have ever taken up speech-writing 
seem children compared to him. If, however, that should 
not satisfy him, it would be no wonder if a divine impulse 
should lead him to higher things still ; for, my dear Phaidros, 
there really is philosophy in the man (279 a). 

It is important not to overlook the dramatic setting here, 
It is Sokrates, not Plato, who pays Isokrates this hand- 
some compliment, and, of course, Sokrates cannot speak 
otherwise than prophetically of anything but the forensic 
speeches of which Isokrates was afterwards ashamed. 
On the other hand, Plato would not have been likely to 
put into the mouth of Sokrates a prophecy that had 
not in some measure been fulfilled. I take it, then, that 
this is a perfectly sincere compliment, and that the tradi- 
tion which represents Plato and Isokrates as friends is 
much more likely to be right than modern speculations 
about a feud between them. They differed, indeed, on 
fundamentals, but they had a good many opinions in 
common, especially about politics. Plato must have 
understood and sympathised with the ideals of Isokrates 
regarding Greek union against Persia, while Isokrates 
would appreciate the Sicilian projects of Plato, which 
we shall have to consider later, though he doubtless 
thought it very absurd of him to begin the training of 
a prince with mathematics. The main point is, however, 
that both Isokrates and Plato were convinced that the 
future of Hellas depended on the revival of monarchy, 


a conviction which the course of history showed to be 
well founded. 

1 66, Where Plato and Isokrates differed was in their 
conception of education. Isokrates was what we call a 
humanist, and the rivalry between him and Plato was 
really the first chapter in the long struggle between 
humanism and science. It must be remembered, how- 
ever, that Greek humanism was of necessity a far shallower 
thing than what we call by the name. In the first place, 
modern humanism has gained immeasurably from having 
to deal with the language and literature of other peoples, 
and especially with those of classical antiquity. An 
exclusive preoccupation with the literature of one's own 
country always tends to shallowness. That is why even 
Roman humanism, as we know it in Cicero, for instance, 
is a far deeper thing than the contemporary Greek 
rhetoric. It has Greek antiquity as well as Roman 
behind it, and that gave it strength. The humanism 
of the Renaissance, again, was saturated with the results 
and spirit of Greek science, and so prepared the way for 
the scientific discoveries of the sixteenth and seventeenth 
centuries, while Greek humanism inherited from the 
Sophists of the fifth century a rooted distrust of science 
and scientific methods. The humanism of Isokrates had, 
therefore, hardly any real content, and tended to become 
little more than the art of expressing commonplaces in a 
perfect form. 

1 67. At the same time, the form invented by Isokrates 
really was perfect in its way, and he has, directly or indi- 
rectly, influenced every writer of prose down to the present 
day. Even commonplace thinking may have its value, 
and it is a very good test of that to express it in an 
artistic way. If one has to utter one's thoughts in ac- 
cordance with a prescribed scheme, they will at least gain 
in lucidity and coherence, so far as they are reasonable at 
all. Thoughts that are wholly unreasonable do not admit 
of artistic expression. In this way Isokrates was quite 
entitled to claim that his teaching was of service to his 


pupils, and he certainly did a great deal to make Hellenism 
a possibility, in spite of the fact that his own political 
thinking is unduly coloured by the rhetorical antithesis of 
Hellenes and barbarians, a division of mankind which 
Plato regarded as unscientific (Polit. 262 d). At any rate, 
whatever we may think of Isokrates, there can be no 
doubt that Plato recognised his merits, and it is curious 
to note how, the more he came to diverge from him on 
matters of greater importance, the more he fell under the 
fascination of his style. It is just in these later dialogues 
where the scientific spirit is most dominant that the 
influence of Isokrates may be traced most clearly. In 
every other respect such a work as the Sophist is wide 
as the poles asunder from anything Isokrates was capable 
of understanding, and yet it is in that very dialogue that 
Plato for the first time troubles to avoid hiatus, and even 
adopts some specially Isokratean devices for doing so. It 
seems as if, when he felt his own gift of artistic writing 
beginning to fail, he was glad to reinforce it in this way. 

1 68. To Plato philosophy was, of course, something 
quite different from what it was to Isokrates. If we look 
at the dialogues he was writing about the time he founded 
the Academy, and especially the Symposium, the Republic, 
and the Phaedrus, we shall see, I think, that he regarded 
it chiefly in two lights. In the first place, it is the con- 
version of a soul, and in the second place it is the service 
of mankind. We shall take the latter point fyrst, because 
it is impossible to understand Plato's object in founding the 
Academy till it has been made clear. No one has insisted 
more than he has on the necessity of disinterested scientific 
study, freed from all merely utilitarian preoccupations, but 
at the same time no one has maintained more firmly that 
such study is only justified in the last resort by the service 
it can render to human life. The Sokratic demand that 
the man who knows shall rule had, he tells us (Ep. vii. 
326 a), taken the more precise form that the only hope 
for mankind is that kings should turn philosophers or that 
philosophers should become kings. That ideal never left 


him, and, though he ceased to hope for its realisation, he 
was always ready to welcome any approach to it. In 
default of the philosopher king much might be effected 
by the co-operation of a philosopher and a tyrant, especially 
if the latter was young and impressionable. He reaffirms 
this conviction in the Laws (709 e), though he had already 
been disappointed in one attempt to work upon that plan. 
The Academy was first and foremost, then, an institution 
for training rulers and legislators, and it was extremely 
successful in its task. It was, in fact, made a charge 
against it that it produced tyrants, which is true enough, 
and much to its credit, if the facts are rightly estimated. 
It also produced Its fair share of tyrannicides. 

Isokrates boasts that his training was more practical than 
that of his rivals, but most of his pupils turned out rheto- 
rical historians or rhetorical tragedians, while Plato trained 
statesmen and men of science. We shall see later that 
the Academy was often applied to for legislators by new 
communities. There is not the slightest improbability in 
the story that Epameinondas, who had been an associate 
of the Pythagorean Lysis, asked Plato himself to frame a 
code of laws for Megalopolis, though we are told that Plato 

The Methods of the Academy. 

169. Two methods are specially associated with Plato's 
name, that of Analysis and that of Division. The former, 
indeed, is said to have been invented by Plato, who 
" delivered it " to Leodamas, and it is significant that in 
Book XIII. of Euclid, which is in a pre-eminent sense the 
work of the Academy, analytical proofs are given for the 
first time in addition to those in the usual form. It can 
hardly be supposed, however, that analysis is no older than 
Plato. The proof called apagogic (reductio ad absurdum) 
is an application of the analytic method, and it was certainly 
used by the Pythagoreans. Moreover, Plato himself repre- 
sents Parmenides as teaching it to Sokrates, while in the 
Meno and Phaedo y as we have seen ( 121), Sokrates himself 


explains it. It follows that what Plato did was at most to 
formulate the method more clearly, and very probably to 
show the necessity of supplementing analysis by synthesis, 
in order to secure that all the intermediate steps discovered 
by the analysis are reciprocal. 1 The chain of consequences 
must be reversible if the proof is to be complete. Each 
analysis given in Euclid is immediately followed by the 
corresponding synthesis. This was revived by Galileo in 
the seventeenth century as a substitute for the prevailing 
Aristotelian methods. 2 

170. The other Platonic method is that of Division 
(Sialpeans), which even the comic poets knew to be charac- 
teristic of the Academy. As analysis aims at explanation 
or proof, so division is the instrument of classification or 
definition. The method is this. The thing to be defined 
or classified is first referred to its genus, and then, by a 
series of dichotomies, the genus is divided into species and 
sub-species. At each division we ask to which of the species 
it gives us the thing to be defined belongs, and that is 
divided once more, the "left-hand" species being left 
undivided as irrelevant to our purpose. The definition 
is found by adding together all the species " on the right- 
hand side." The examples of this method which Plato 
gives in the Sophist and Statesman are only to be understood 
as more or less popular and playful applications of it, but 
just for that reason they serve to show what is meant better 
than a serious example, where it would have been necessary 
to justify each step elaborately. We shall return to this 
subject when we come to the Phikbus. 

171. As to the plan of teaching and study adopted 
in the Academy we have, as is natural, but little direct 
evidence, but what we have is at once trustworthy and 
instructive. In the first place, there can be no doubt that 
Plato gave regular lectures (crvvovviai, a/cpoaoW), and that 

1 This was the view of Tannery. 

2 The metodo risolutlvo is just the avaXvriKr) /*#oSos, Galileo was a 
convinced Platonist. 


his hearers took notes. Aristoxenos said that Aristotle 
" was always telling " how most of those who heard the 
lecture on the Good were affected. They came expecting 
to hear about some of the recognised good things, and 
when they heard of nothing but Arithmetic and Astronomy 
and the Limit and the One, they thought it all very strange. 
We know from Simplicius that Aristotle, Speusippos, and 
Xenokrates had all published their notes of this very dis- 
course. We may infer that Plato did not write his lectures, 
and that is confirmed by Aristotle's reference to his <c un- 
written dogmas " (aypa<pa Soy^ara). As we know, Plato 
did not believe in books for serious purposes. In the 
Seventh Epistle he complains that, even in his lifetime, 
some of his hearers had published accounts of his doctrine 
of the Good, which, however, he repudiates. The passage 
is worth quoting. He says : 

There is no writing of mine on this subject, nor ever shall 
be. It is not capable of expression like other branches of 
study $ but, as the result of long intercourse and a common life 
spent upon the thing, a light is suddenly kindled as from a 
leaping spark, and when it has reached the soul, it thence- 
forward finds nutriment for itself. I know this, at any rate, 
that if these things were to be written down or stated at all, 
they would be better stated by myself than by others, and I 
know too that I should be the person to suffer most from 
their being badly set down in writing. If I thought they 
could be adequately written down and stated to the world, 
what finer occupation could I have had in life than to write 
what would be of great service to mankind, and to reveal 
Nature in the light of day to all men ? But I do not even 
think the effort to attain this a good thing for men, except for 
the very few who can be enabled to discover these things 
themselves by means of a brief indication. The rest it would 
either fill with contempt in a manner by no means pleasing 
or with a lofty and vain presumption as though they had 
learnt something grand (341 c-e). 

This is not mystery-mongering, as has been said ; it is 
simply a statement of the true theory of all higher educa- 
tion. To be of any use, philosophy must be a man's very 


own ; it ceases to be philosophy if it is merely an echo of 
another's thought The passage is also a salutary warning 
to the interpreter of Plato. He may, in a measure, re- 
cover the dry bones of his deepest thought ; the spirit of 
it is less easy to reproduce. 

172. We are to think, then, of Plato lecturing in the 
Academy without notes, and of his more attentive hearers 
taking down what they could. But the set discourse, 
though necessary, was by no means the most important 
part of the work. It was better than a book, no doubt, 
but it was only preparatory to the real thing. Its function 
is to rouse the soul, to turn it to the light, but the soul 
must see the light for itself. The Academy was no mere 
lecture-hall ; it was an institute for scientific research. 
Simplicius, who had the library of the school at his dis- 
posal, tells us that Plato, who held that the movements 
of the heavenly bodies must be regular, cc propounded 
it as a problem " to the mathematicians of the Academy 
to find on what hypothesis (TLVMV uTrore^eVrcoz/) their 
apparent irregularity could be explained so as to "save 
the appearances." 1 The word c< problem" calls for special 
attention in this connexion. Both it and " protasis," 
the verb corresponding to which (irpoTeivGiv) has been 
rendered " propound " (proponere) in the passage just 
referred to, originate in the Greek custom of asking 
riddles at banquets, and the convivial associations of the 
words bear witness to the idea of scientific research as a 
common life (TO o-v^Jv). That accounts in turn for in- 
vestigation taking the form of a quest for solutions (\vo-eti) 
of certain problems (V^o/SX^ara) or difficulties (aTroplai). 
We have a collection of such in the Aristotelian corpus, 
which is obviously derived from the work of his school, 
and the passage of Simplicius just quoted shows that the 
method originated in the Academy. It is, of course, the 
beginning of the system of education through original 

It is to be observed further that Plato by no means 
1 Simpl. de Caelo, pp. 488. 21 ; 4.92. 31 (Heiberg). 


confined the researches of his students to subjects of 
special interest to himself, such as mathematics and 
astronomy. No doubt they had all to go through a pre- 
liminary course of mathematical training, but there is 
abundant evidence that biological studies were also pursued 
with enthusiasm. The satire of the comic poets was 
largely directed to this side of the Academy's activity. 
Epikrates (fr. 5) laughs at Plato, Speusippos and Mene- 
demos for investigating by the method of division to what 
genus the pumpkin belongs. Speusippos, Plato's nephew 
and successor, wrote many books on the classification 
of animals and vegetables, and the few fragments that 
remain deal, for instance, with shell-fish and fungi. In 
the Critias (no d sqq^ Plato himself surprises us by an 
account of the geological history of Attika and its 
economic consequences which is almost on a level with the 
most modern discussions of the kind. The biological 
work of Aristotle belongs to the early period of his life, 
and it is natural to bring that into connexion with these 
facts. It remains to be said that we must of course 
represent the Academy to ourselves as well provided with 
scientific apparatus and collections. Aristophanes takes 
it for granted in the Clouds that a scientific school would 
possess maps and astronomical models as a matter of 
course, and, if that was so in the fifth century, it may 
certainly be assumed in the fourth. 

The Programme of Studies. 

173. We may fairly take the higher education of the 
Guardians outlined in the Republic as a guide to the course 
of study followed in the Academy. We are expressly told 
that the mathematical part of the course is to occupy the 
ten years from twenty to thirty, and it has all the appear- 
ance of a regular programme. It would, however, be a 
mistake to suppose that what is said about the sciences in 
the Republic represents the mature thought of Plato on the 
subject. It was written either before the foundation of 


the Academy or very shortly after, and the theories most 
characteristic of Plato's teaching are not yet elaborated. 
He is quite conscious of that. What he proposed was a 
thorough criticism of the hypotheses of all the sciences, 
and that had not yet been carried out. That is what he 
means by the " longer way," which has yet to be travelled 
(435 d, 504 b). We must be prepared to find, then, that 
in some important respects the philosophy of the exact 
sciences given in the Republic is completely transformed at 
a later date. 

The programme is based on the principle that the 
function of education is the conversion (Trepurrpocfrfy of 
the soul from the contemplation of Becoming (<yeVe<w) to 
that of Being (ova-la). As we have seen, that distinction is 
Pythagorean, and it is therefore natural that the course 
should consist of the four Pythagorean sciences which sur- 
vived in the medieval quadrivium, though with this dis- 
tinction, that plane and solid geometry are distinguished, 
so as to give five studies (yuaft^ara) instead of four. If 
we take these in order, we shall see the point of view from 
which Plato started. 

I. Arithmetic. At this stage, Arithmetic is to be 
studied, not for utilitarian or commercial purposes, but 
with a view to understanding the nature of numbers by 
thought alone. It arises from the ambiguity and relativity 
of sense perception. What appears one to the senses also 
appears as many from another point of view. Two appear 
as one and one as two, so it is the function of thought to 
distinguish and separate these from the confusion in which 
they are presented by sense. It is the business of Arith- 
metic to consider numbers by themselves, not visible or 
corporeal numbers. A visible or tangible unit admits of 
division, and so is many as well as one, but unity itself is 
indivisible. Visible and tangible units are not necessarily 
equal to one another, but the units of the arithmetician 
are all absolutely equal. Such units cannot be apprehended 
by sense, but only by thought, and that is what gives the 
study of arithmetic its educational value (524 b 526 c). 


2. Plane Geometry. Geometry too is to be studied for 
other than utilitarian ends, for which) indeed, a very slight 
knowledge of it is required. Though geometers talk of 
performing certain operations, such as " squaring " and 
"applying" and "producing," that is only a manner of 
speaking, and Geometry too has to do with Being, not 
with Becoming. Its objects are certain spatial relations 
which simply are> whatever we may do, and do not come 
into being in virtue of our constructions. This study too, 
then, is of value as purifying an instrument of the soul 
(527 a-e). 

3. Solid Geometry. Sokrates is about to pass from 
Geometry to Astronomy, but recollects himself and points 
out that there is a science intermediate between them, that 
which deals with the cc third increase " (T/>/TJ/ a?^), that is, 
with the cube, and generally what has three dimensions, 
depth as well as length and breadth. "But," says Glaukon, 
" that does not appear to have been invented yet." 
Sokrates answers that this is because in the first place no 
state holds such studies in honour, and in the second, 
because a director (eTna-rar^i) is required to guide them. 
If the state were to second the efforts of such a director, 
they would soon be perfected. Even as it is, their extreme 
elegance (x<*jw, T ^X / 04 ) causes them to make some 
progress (528 d). 

As has already been indicated, this remarkable passage 
appears to refer to the fact that, though the Pythagoreans 
had made a beginning, the theory of the five regular solids 
was completed for the first time by Theaitetos, while the 
problem of the duplication of the cube was not solved till 
a still later date. The term Stereometry is not used here ; 
it appears for the first time in the Epinomis (990 d). 

174. The remaining studies deal with motion, and it 
is hinted that there may be more than the two mentioned. 

4. Astronomy. Astronomy is not to be studied merely 
for its use in agriculture, navigation, or strategy, or even 
because it turns our eyes upwards to a higher world. The 
visible motions of the heavenly bodies with all their 


labyrinthine intricacy are related to true astronomy only 
as the diagrams analysed by the geometer are related to 
his science, that is to say, these apparent motions must be 
regarded merely as illustrations (xa^a^e/y/uara). We must 
treat them as " problems" (TrpofSXtfimacriv xjow/xei/o*), not as 
solutions. What we have to study is " the true motions 
with which the real velocity and the real slowness move in 
relation to one another, in the true numbers and the true 
forms, and carry their contents with them" (529 d). 

This sentence is easily misunderstood and requires 
elucidation. In the first place, the visible motions of the 
heavenly bodies are what we call their apparent motions, 
which are of great complexity and at first sight seem quite 
irregular. The planets move at one time from east to 
west among the stars, at another from west to east, and 
sometimes they are stationary altogether. That is the 
" problem" we have to solve. The " real velocity" (TO oj/ 
raxps) is spoken of simply as opposed to the apparent 
velocity. We should not think it necessary to add " the 
real slowness/' but that is only an instance of the Greek 
tendency to "polar expression," and has no serious im- 
portance. We may speak of a lesser velocity as a " slow- 
ness" if we please. Then this velocity is spoken of as 
carrying its "contents" (ra ei/oVra) with it. That is 
because the Greeks were in the habit of attributing the 
orbital revolution to the orbit itself, and not to the celestial 
body, which was regarded as occupying a fixed place in its 
orbit. That again is due to their regarding all orbital 
revolution as similar to that of the moon, the only case 
which can be adequately studied without a telescope. 
The moon always presents the same face to the earth (or 
nearly so), and, in the absence of any indication to the 
contrary, it was not unreasonable to suppose the other 
planets did the same. We say the rotation of the moon 
upon its axis takes the same time as its revolution round 
the earth ; the Greeks expressed the same fact by saying 
the moon does not revolve at all relatively to its orbit. 
That is why Aristotle can urge the fact of the moon's 


always presenting the same face to us in support of the 
view that none of the heavenly bodies rotate. To us that 
is just what proves the moon does revolve on its axis, but 
Aristotle is thinking of the orbit (or rather, in his case, the 
sphere) to which the moon is attached. All this explains 
why it was natural to speak of the heavenly bodies as the 
things "in the velocity" (ei/oVra, sc. 777 ra^vr^ri). 1 The 
"true numbers" are the number of days and years the 
revolutions take, and the "true forms" are the circles, 
spirals, or whatever they may prove to be, which they 
trace. What is meant, then, is simply that we must have 
a science which will exhibit the true motions of the heavenly 
bodies and not the motions they appear to have. The 
apparent motions of the heavenly bodies no more express 
the laws of solid bodies in motion than the diagrams of the 
geometer embody the truths of geometry. 

It is amusing to observe that such a utilitarian thing as 
"Greenwich time'* has to take account of this. Our 
watches are set, not by the visible sun, but by an " intelli- 
gible" sun called the "mean sun," which only coincides 
with the visible sun four times a year, and then only for an 
instant. That this illustration is not too far-fetched is 
shown by the fact that the apparent anomaly of the sun's 
annual course was just one of the problems we know to 
have been investigated in the Academy. 2 It may be added 
that this is fatal to the interpretation which makes Plato's 
astronomy refer to some imaginary " ideal" heavens. If 
it had, why should he have troubled himself about the 
sun's anomaly ? It would have been so easy to say that 
the intelligible sun had a uniform velocity, and to disregard 
the shortcomings of the visible sun. 

5. Harmonics. The next study is Harmonics, which 
the Pythagoreans regard as the counterpart of Astronomy. 
As the one deals with motions apprehended by the eye, so 
does the other deal with motions apprehended by the ear. 

1 Adam's interpretation of this passage is sufficiently refuted by the 
fantastic account he has to give of TO, li/ovra. 
* Simplicius in Phys. p. 292. 22 (Diels). 


The same principles will apply here. Not to speak of 
those who attempt to determine the harmonic intervals by 
ear, even the Pythagoreans themselves, who express them 
by numerical ratios, do not sufficiently emancipate them- 
selves from the sound as heard. 1 It is not enough to say 
that such and such an interval is expressed by such and 
such a ratio ; we ought to consider which numbers are 
consonant with one another and which are not, and to ask 
the reason of this in both cases. 

Here, as in the case of Astronomy, we have an anticipa- 
tion of the science of a later age. The sounds we hear are 
produced by a succession of "beats" (^X^a/) of the air 
(we should say, of waves), and the business of the musical 
theorist is to express the differences of the musical intervals 
in terms of these, and not merely in terms of the length of 
strings. So far as the Pythagorean system goes, it would 
seem that the consonances might be expressed by any other 
ratios just as well as those which have been experimentally 
discovered. In fact, the Pythagorean intervals are a 
problem and not a solution. The fact that some intervals 
are consonant, while others are not, must be due to some- 
thing in the nature of number itself. 

175. All these studies, however, are but the prelude 
to the strain we have really to learn, and that is Dialectic. 
We know already what Dialectic means in the Sokratic 
sense. It is the art of question and answer, the art of 
giving a rational account of things and of receiving such an 
account from others (StSovai KOL Se^eo-Gai Xoyoi/). Even 
Xenophon knew that Sokrates made those who associated 
with him "dialectical," though he attributes to him an 
erroneous etymology of the word. 2 But here something 
more is meant than the art of reasoning, or at any rate some- 

1 Aristoxenos represents the first class for us and Archy tas the second. 

2 Mm. 'iv. 5. 12. He makes him derive the verb SiaXeytcrOai from 
StaAeyctv Kara ykvt] TO, Trpay^ara. That is just like the derivation of 
cro^tcrr^s from o rcov cro<<3i/ icrrr/s ( = eTrtcrr^cov) in Prof. 312 c or that 
of vTToQecri's from vTTOTiOrjfJu, " lay a foundation," implied in Re-p. 5 1 1 b. 
The Cratylus is full of such things, so Sokrates may really have said it. 


thing more special. In the Euthydemus (290 c) we are told 
that arithmeticians, geometers 3 and astronomers must hand 
over their discoveries to the dialectician for examination. 
Here we learn (533 b) that the weakness of the method of 
hypothesis, as described for instance by Sokrates in the 
Phaedo^ is just this, that the hypothesis itself is only esta- 
blished by the consistency of its consequences ; it has not 
itself been examined in the light of any higher principle. 
We are told, accordingly, that, though geometers and the 
rest do in part attain reality, they only see it <c in a dream." 
So long as they use hypotheses and refuse to let them be 
moved, because they can give no account of them, they 
cannot be said to behold true Being with a waking vision. 
If we take for our starting-point what we do not know, 
and our end and all the intermediate steps are only a con- 
catenation (crvyu-TrAo/c??) of what we do not know, that is a 
mere agreement (ojuioXoyia) not to raise ultimate questions, 
and cannot become science in the true sense of the word. 

The defect of the special sciences is, then, that they 
depend on hypotheses of which they can give no account, 
and are therefore obliged to use sensible diagrams. We 
are told quite distinctly that Dialectic proceeds by 
"destroying the hypotheses" (avcupova-a ra? uTro^e'aw). 
This has given much trouble to some interpreters, who 
find it hard to believe that Plato desired, for instance, to 
"destroy" the hypothesis of three kinds of angles, which 
he expressly mentions in this connexion (510 c) as funda- 
mental in geometry. It is impossible, however, to take 
the word I have rendered cc destroy" (avatpeiv, to Here) other- 
wise ; for we have seen ( 125) that it is a technical term in 
this context. Further, the view of science taken in the 
Republic really does demand the destruction of the hypo- 
theses of the special sciences. The hypothesis of three 
kinds of angles has a spatial character, and that is just why 
the geometer is forced to use sensible diagrams. The 
ideal is that Arithmetic, Geometry, and the rest should all 
be reduced to one science, and this cannot be done so long 
as their special hypotheses remain. It is only when these 


have been removed that we can ascend to a first principle 
which is no longer a postulate (to an awn-dOeros />x>?), 
namely, the Form of the Good. Then, and not till then, 
can we descend once more without making use of sensible 
diagrams of any kind. The whole of science would thus 
be reduced to a sort of teleological algebra. 

Eukleides and Plato. 

176. We shall understand this point of view better if 
we consider how natural it was that, when Plato set him- 
self to draw up a scheme of scientific study for the 
Academy, he should be influenced by the teaching of 
Eukleides of Megara. He had taken refuge with him 
after the death of Sokrates, and the prominence given 
to Phaidon as the narrator of the last discussion of 
Sokrates on earth points in the same direction, for the 
school of Elis founded by him was closely related to that 
of Megara. Plato was also influenced, of course, by the 
Pythagorean associates of Sokrates, but it looks as if he 
did not become personally intimate with the leading 
Pythagoreans of his day till later. He would have little 
time for that during his first visit to Italy and Sicily. 
This makes it necessary for us to learn all we can about 
Eukleides. It is not much, unfortunately, but the few 
statements we have rest on the best authority, and are 
of fundamental importance. 

In the first place, as we have seen already (117), 
Eukleides was an Eleatic, and the doctrines of the Megaric 
school in a later generation, as we know them from 
Aristokles, 1 still bear traces of their Eleatic origin. 
Accordingly, though we are not entitled to ascribe all 
these doctrines to Eukleides himself without more ado, 
we cannot go far wrong in crediting him with those that 
are definitely Eleatic in character. To begin with, we are 
told that the Megarics considered it their business to 

1 Aristokles was the teacher of Alexander of Aphrodisias. The state- 
ments referred to are preserved in Enseb, Pr, Ev. xiv. 17. 


"throw" (/cara/SaXXeij/) l sensations and appearances and 
to trust to reasoning alone. That goes without saying 
in an Eleatic. We are also told that they held that' 
Being was one and the Other is not, and that there 
was no such thing as coming into being or ceasing 
to be or motion. That is also sound Eleatic doctrine, 
and may be confidently attributed to Eukleides. It 
is impossible, then, to suppose that he could have 
accepted, and still less that he could have originated, 
the doctrine Plato attributes to Sokrates in the Phaedo, 
for there we have a plurality of forms which enter 
into the world of becoming. Eukleides accordingly, 
though present, takes no part in the discussion. On the 
other hand, he appears to have been deeply interested in 
the teaching of Sokrates on the subject of the Good. 
We still have a curious document written in the Doric 
dialect, in which certain Sokratic doctrines about good- 
ness are clearly referred to. 2 It is generally recognised 
that it belongs to the end of the fifth century, and its 
" eristic " character, taken in conjunction with its Doric 
dialect, strongly suggest Megara as its place of origin. 
At any rate, we know that Eukleides identified the Good 
with the One, which is also called by other names, such as 
God or Wisdom. It is only possible to guess his exact 
meaning, but the fact of the identification is certain, and 
its connexion with the teaching of Sokrates seems plain. 
As there is nothing else than the One, he inferred that 
there is no such thing as evil. The method by which it 
is shown that the senses and the things that appear to 
them are unreal, is to show that there are "two state- 
ments" (Sicro-oi Xo'yot) which may be made with equal 
truth and cogency about all of them. That is what the 
Megarics called Dialectic and their opponents called Eristic. 
If we 'may trust Aristotle's account of the matter, the 

1 See p. 113, . 2. 

2 The Sio-o-ot Xoyot (formerly known as Dialexeis). It is printed in 
Diels, Fors? ii. pp. 334 sqq. See Taylor, Tana Socratica, i. pp. 91 sgg. 


method had degenerated by his time into a mere quibbling 
about words. It does not follow that it was anything 
but a serious doctrine in the hands of Eukleides ; for Plato 
had not yet cleared up the meaning of " is " and " is not," 
and we shall see good grounds for believing it was just 
his interest in the teaching of Eukleides that led him to 
do so. It is highly probable, then, that the account of 
Dialectic in the Republic was written under this influence, 
and in that case we can most easily understand it as an 
effort to do justice to the position of Eukleides without 
following him in reducing all the forms to the intelligible 
One, which is also somehow the Good. I have said (129) 
that I regard the doctrine of the Good as Sokratic, but 
there are some things said about it in the Republic which 
seem to be Plato's own, for they are directed against 
the identification of the form of Good with Being on 
the one hand and Wisdom on the other, and these are the 
doctrines of Eukleides. According to the Republic, the 
Good is neither Being nor Knowledge, but the cause of 
both. It altogether transcends and is " on the other side " 
of Being (eTre/eewa rfjs over/a?), as it transcends Knowledge. 
In some such way as this, it may have seemed to Plato at 
the time, the monism of Eukleides might be avoided, 
while all that was valuable in his system might be 

The theory which would naturally follow from this way 
of regarding the Good would be one of "emanation," 
and that is in fact the view which was associated with it 
when the doctrine was revived in later days. To a con- 
siderable extent Neoplatonism may be fairly described as 
a development of the thought that was in Plato's mind 
when he wrote this part of the Republic. We have no 
means of knowing how far Plato himself had gone in this 
direction. He could not in any case have made Sokrates 
the mouthpiece of such a theory ; and 5 as has been indi- 
cated, he has probably strained historical verisimilitude to 
some extent in saying as much as he does. We shall 
never know more on the subject, for he never speaks 


in this way of the form of Good again, and Aristotle 
never even alludes to this passage. As we shall see, the 
solution that finally commended itself to Plato was reached 
on other lines, and we have now to consider the steps by 
which he finally emancipated himself from the Megaric 



177. Plato's emancipation from the influence of 
Eukleides seems to have been gradual. For about 
twenty years he carried on his work in the Academy 
without interruption, and it does not appear that he 
published any more dialogues till towards the end of 
that period. His hands were probably too fulL A time 
came, however, when he felt it necessary to define his 
attitude to other philosophers, and that could only be 
done by writings addressed to a wider circle than the 
school. We cannot estimate the interval of time which 
separates the Theaetetus from the Republic and the Phaedrus, 
but it was probably one of a good many years. When 
Plato began to write dialogues again they had a different 
character from those of his early life. This is marked 
first of all by a significant change in form. Some of the 
very earliest dialogues had been simple dramatic sketches 
in direct speech, but this form soon proved inadequate for 
Plato's purpose, so long as that was mainly to give a 
picture of Sokrates as he lived and moved. Unless 
interpreted by action it makes too great a demand on the 
reader, who has to supply the mise en sdne and the stage 
directions himself. Narrated dialogue, on the other hand, 
allows of descriptions and comments which make the 
picture live, and all the most artistic of Plato's dialogues 
are therefore narrated. When, however, the scientific 
interest begins to prevail over the artistic, this form 
becomes very cumbrous. We see it at its worst in the 


Parmenides, the formula of which is cc Antiphon said that 
Pythodoros said that Parmenides said." In the Theaetetus 
there is an express reference to this question of form. 
Like the Phaedo and the Parmenides^ that dialogue opens 
with a short dramatic introduction ; but this leads up, not 
to a narrated dialogue as in their case, but to one 
which is also dramatic in form. That, we are told . 
(143 c)> is to avoid the troublesome repetition of such 
phrases as "And I said," "He assented/' He agreed." 
It is true that the Parmenides is probably a little later than 
the Theaetetus, but they both belong to the same period, 
and Plato may well have been engaged on the one when 
he produced the other. If so, we can easily understand 
his conceiving a distaste for the narrative form. At any 
rate, he never made use of it again, and his latest dialogues 
are simply dramatic, just as his earliest had been. 

178. Philosophically, the distinguishing feature of 
these dialogues is Plato's preoccupation with the Megarics. 
The Theaetetus is dedicated to Eukleides, or rather to his 
memory ; for it is not likely that he was still living. 
Plato does not introduce living characters if he can 
help it. He was about to criticise the doctrine of 
Eukleides, and the Theaetetus is meant to lead up to that 
criticism, but he still cherished, we may suppose, a feeling 
of regard for the man. Nor is there anything in the 
dialogue that directly impugns his doctrine. It does not, 
we shall see, go far beyond the possibilities of discussion 
within the Sokratic society itself. The rift, as has been 
pointed out ( 129), was probably in existence before the 
death of Sokrates, but was regarded as a difference within 
the school. For the same reason, there is no difficulty 
in making Sokrates the chief speaker. And yet the point 
of view is no longer strictly Sokratic. Plato is now as much 
impressed by the dangers of a one-sided intellectualism as 
by those of a one-sided sensationalism. He avoids the 
doctrine of forms altogether in this dialogue, though there 
are points in the argument where we should expect it to be 
discussed. It was taking another shape in his mind by 


this time, and he could not make Sokrates the mouthpiece 
of that. 

179. This brings us face to face with the very im- 
portant question of the place assigned to Sokrates in the 
dialogues of Plato's maturity. The discussion narrated in 
the Theaetetus is supposed to have been taken down by 
Eukleides and revised and corrected by Sokrates himself 
(143 a). Further, it is supposed to be read aloud at 
Megara years after the death of Sokrates. The informal 
discussion of the earlier dialogues has become a deliberate 
statement of doctrine intended to be read and criticised. 
As, however, it only states a problem which had really 
been raised by Sokrates, and does not give the solution, 
there is no difficulty in his being the chief speaker, though 
by a curious device, certain doctrines are said to have been 
known to him only " in a dream." The Parmenides is also 
represented as a deliberate statement ; for it is supposed 
to have been learnt by heart and repeated long afterwards, 
a fiction which would seem more credible then than in this 
age of books. This dialogue contains a direct criticism of 
the doctrine of forms as that is stated in the Phaedo and 
the Republic^ and the introduction of Parmenides as the 
chief speaker suggests that it was the Eleatic criticism that 
in fact forced Plato to seek for a more satisfactory formula- 
tion of it. He was bound to make his position clear ; for, 
whether he himself had ever held the doctrine criticised or 
not, he had certainly done a great deal to propagate it by 
his Sokratic writings. Clearly Sokrates cannot be the chief 
speaker here, but it would have been unseemly to introduce 
Eukleides, for instance, as criticising him. So Plato takes 
advantage of the visit of Parmenides and Zeno to Athens 
almost a century before to put the criticism into the mouth 
of the founder of the school to which Eukleides belonged. 
It would have been too much, however, to represent Par- 
menides as asserting the reality of " not being," which is 
the theme of the Sophist, so the leading part in that dia- 
logue and its sequel, the Statesman^ is taken by an Eleatic 
stranger, who is a very unorthodox disciple of the great 


Parmenidcs. Plato seems to mean by introducing this 
enigmatic figure, who certainly expresses his own views, 
that he himself, rather than the disciples of Eukleides, was 
the true successor of Parmenides. In the Philebus we seem 
to come nearer Plato's own philosophy than we do any- 
where else, and yet Sokrates is once more the chief speaker. 
That is a problem we shall have to face later. In the Timaeus 
and Critias Sokrates is only a listener, and in the Laws he 
does not appear at all. We are told in the Phaedo that 
Sokrates had rejected all attempts at a mechanical explana- 
tion of the world, and the Timaeus contains such an attempt. 
As to the works which deal with human history and insti- 
tutions, like the Critias and the Laws., we learn from the 
Timaeus (i 9 a-d) why Sokrates can take no part. He could 
paint the picture of an ideal state, but he could not make 
the figures move. He is made to confess that he could 
not, for instance, represent his state as engaged in the 
struggle for existence with other states ; to do that men 
are required who by nature and training have a gift for 
practical politics as well as for philosophy. This is a very 
valuable passage as evidence that Plato was conscious that 
some themes were appropriate for Sokrates and others 
were not. The implied criticism of his master's political 
teaching should also be noted. Plato knew very well 
that, on its constructive side, it was too uncompromising 
and on its critical side too negative. That is partly 
why so many followers of Sokrates turned out reactionaries 
rather than statesmen. 

The Theaeteius. 

1 80. The purpose of the Theaetetus is to clear the 
ground by showing that knowledge cannot be identified 
either with sensation or with thought Theaitetos, after 
whom the dialogue is named, was one of the original mem- 
bers of the Academy and one of the most distinguished, 
and we gather that he died of wounds and dysentery after 
a battle at Corinth, which was probably that of 369 B.C. 


It was certainly before this dialogue was written ; for the 
beautiful description of his character in the introduction 
can only be read as a tribute to a gifted disciple too soon 
lost His eminence as a mathematician is skilfully sug- 
gested by the story of how, when a mere lad, he discovered 
a general formula for numbers of which the square root is 
irrational. It seems probable that his death was still recent 
when the dialogue was composed, and for that and other 
reasons it is most probably dated in 368 B.C. or a little 
later, when Plato was about sixty years old. The other 
speakers are the " younger Sokrates," the friend of Theai- 
tetos, and like him an original member of the Academy, 
and the mathematician Theodoros of Kyrene. He had 
been a follower of Protagoras and a friend of Sokrates. 
He therefore belongs to an earlier generation than the two 
lads whose teacher he is, and had certainly passed away 
long before this dialogue was written. The dialogue is 
supposed to take place just before the trial of Sokrates 
(210 d), that is to say, more than thirty years before it was 

1 8 1 . The first serious answer given by Theaitetos to 
the question, "What is knowledge ?" is that it is sensation 
(ala-Oqcrii). That definition agrees with what Protagoras 
said in another form about knowledge, namely, that man 
is the measure of all things, of what is that it is, and of 
what is not that it is not. This means that as a thing 
appears to me, so it is to me, and as it appears to you, so 
it is to you. Instead of saying "as a thing appears to me," 
we may equally well say "as I am sensible of it," for 
instance," A wind appears to me cold" is the same thing 
as "I am sensible that a wind is cold." In a word, 
appearance (fyavTavia) and sense (ar0?(w) are the same 
thing in the case of hot and cold and the like. Sensation, 
then, is always sensation of what is, and cannot err ; for 
what is is that of which I am sensible (152 a-c). 

That, however, was only a dark saying of Protagoras 
addressed to the vulgar crowd ; to the initiated he told 
the truth, and the truth is this. It is not true to say 


tat what appears is. In reality nothing is, everything is 
scorning, as Herakleitos and others have taught. Motion 
the cause of growth, while rest is the cause of decay and 
casing to be. Motion is good, and rest is evil. You 
innot rightly use the terms " something/' "such a thing," 
one/' ct is"; for, if you say " Something is great," it will 
3pear small from another point of view> and so with the 
ist (152 d). 

In the light of this principle let us consider the case of sight. 
When we use the words" white colour/' we must not suppose 
that what we mean by these words is either something outside 
the eyes or something in the eyes. We must not suppose it 
to be in any place at all. We must say rather that it results 
from the impact (Trpocr/SoA^) of the eye on the appropriate 
movement (irpos TT\V Trpoa-faovcrav (j>opdv) outside it, being 
neither what impinges nor what is impinged upon, but a some- 
thing between the two having a proper character of its own 
for each individual (154 a). Thus no one knows whether 
what appears to him is the same as what appears to another, 
and everyone knows that what appears to himself in one way 
at one time appears to him differently at another. And so 
with other objects; for instance that which after measurement 
and comparison we call great, that which after touching we 
call hot, become respectively small and cold by the presence 
of greater or hotter objects. Six dice compared with four are 
"more" and " half as many again"; compared with twelve, 
they are " less" and " half," yet they are not changed in them- 
selves. They become more and less, and yet nothing has been 
added to them or subtracted from them (153 d 154 d) 

On the other hand, if we look into our own thought, we 
shall agree in the three following propositions : (i) Nothing can 
become greater or less either in size or number so long as it 
is equal to itself; (2) Nothing can increase or decrease to 
which nothing is added or from which nothing is taken 
away ; (3) Nothing can be what it was not before without 
becoming and having become. But all these propositions are 
in direct contradiction to the instance of the dice which we 
considered above, or again to such a case as this " I, Sokrates, 
am now taller than you, Theaitetos ; in a year, I shall be 
smaller (for Theaitetos is still a growing lad), though nothing 
will have been taken from me, nor shall I have become, though 
I shall be, what I was not before" (154 d 155 c). 


Let us go deeper into the mysteries of those wise men 
of whom we spoke, taking care that none of the unini- 
tiated hear us, the"hammer-and-tongs persons" (avrirvTroi 
avdpwTToi), who think that nothing is but what they can 
clutch in their hands, and refuse the right of being to 
actions and processes and everything invisible. The hidden 
truth is this. Nothing is but motion, but there are two forms 
(elStf) of motion, either of infinite extent, the one having 
the power of acting, the other of being acted upon. The 
mutual intercourse of these motions begets an infinity of 
offspring (e/c-yoj/a), each of which is a twin, being partly 
sensation and partly the sensible, the one always simul- 
taneously accompanying the other. Of the infinity of 
sensations many have received names, warming and cool- 
ing, sight, hearing and smell, pleasure and pain, desire and 
fear, and so forth. The corresponding sensible things are 
colours, sounds, and so forth. These motions are quick 
and slow ; those that are slow take place in one spot and 
in relation to what is in contact with them, and are thus 
the producers ; those that are produced are swifter, for 
their motion is from place to place (155 d 156 d). 

Thus what we call seeing may be analysed as follows. On 
the one side there must be the eye, on the other something 
commensurable (<rvfj.iJ.eTpov) with the eye. These are the 
<e slower motions" which take place in one spot. If they 
come into one another's presence, from the former to the 
latter there is a motion, sight ; from the latter to the former 
there is a motion, whiteness. These are the "swifter 
motions " which pass from place to place. This whiteness 
cannot be said to be anything ; it is continually becoming as a 
result of motion. Nor can we even say that what acts or 
what is acted upon is anything that can be fixed and 
individualised in thought ; for the one is not until it meets 
the other, and the one in one combination appears as the 
other in another combination (156 d 157 a). 

Strictly speaking, then, we must not admit any terms such 
as "this," " that,'' "something," but must think of every- 
thing as a process of becoming, being destroyed, being 
changed, and this both in the case of particular sensible 


qualities and of aggregates (aOpola-juiaTa) of particular 
sensible qualities, such as what we call "man/' "stone," 
and every individual object (157 c). 

It only remains to consider the question of the sensa- 
tions of dreaming, insane and diseased persons. We can- 
not prove that what we call dreaming is not waking, and 
vice versa ; for in both states the soul upholds the truth of 
what appears to it at the moment, and so in the case of 
insanity and disease, except that these states last longer 
than sleep. The answer is simple. Sokrates awake or in 
health is, taken as a whole, other than Sokrates in sickness 
or asleep. Accordingly, any natural agent will act upon 
him otherwise in these different states, and the resultant 
of the agent and what it acts on will be different. Now 
the resultant is what it is, not in itself, nor relatively to 
the agent only, nor relatively to Sokrates only, but rela- 
tively to both. When someone becomes sensible, he 
becomes sensible of something, and, when something be- 
comes sensible, it becomes sensible to someone, and what 
the person is or becomes, he is or becomes relatively to that 
thing, and so with the thing. The being or reality (oi3o-/a), 
then, of the moment (i.e. the coexistent, correlative sensa- 
tion and sensible) is bound to both the agents of which it 
is the resultant ; and, from the side of the person, sensa- 
tion, the momentary state, is true ; for it is a sensation of 
what the person at the moment is (157 e 160 d). 

182. This is obviously a well-thought-out and co- 
herent theory of sensation. We are not told whose it was, 
though it is made quite plain that it was not to be found 
in the book of Protagoras ( 92). There are certain 
points in it which remind us of what we are told about the 
Herakleitean Kratylos, who criticised his master for saying 
that we cannot step twice into the same river. We cannot 
do so even once. And yet, if the theory just expounded 
were his, we should surely hear a great deal more about 
him than we do. On the other hand, it can hardly be an 
improvised fiction ; it is too strongly characterised and too 
personal for that. It is, of course, quite on the lines of 



the view of sensation everywhere attributed to Sokrates, 
so there is no difficulty in putting it into his mouth ; but 
it must clearly have been worked out by someone who 
believed in it as an adequate account of knowledge. On 
the whole, it seems best to regard it as in this form Plato's 
own. Aristotle tells us that in his youth Plato had been 
familiar with the doctrine of Kratylos, and had adopted 
it, 1 and there is an earlier dialogue called by the name of 
that thinker, in which Herakleitean doctrine is discussed. 
Aristotle further tells us that Plato continued to hold this 
doctrine to the end, and there is certainly nothing in it, as 
an account of sensation, that he need ever have wished to 
retract. In fact, a thorough-going sensationalism is the 
necessary foundation of Platonism. I assume, then, that 
the doctrine is that of Kratylos, while the elaboration of it 
is Plato's. That will account for the obvious zest with 
which he expounds it, and his equally obvious annoyance 
at the cheap objections which may so easily be made to it. 
These objections are certainly captious enough, and 
Sokrates himself protests that it is treating Protagoras 
unfairly to urge them. He even undertakes to reply to 
them in the name of Protagoras, since he himself is dead. 
They have a certain historical interest ; for some of them 
reappear in the eristic of the later Megaric school, and that 
of itself suggests they may have originated in the circle of 
Eukleides. To discuss them here would merely divert 
the reader's attention from the main argument. As 
Sokrates says (165 d), there is no end to the attacks which 
might be made on the senses by one of these c< mercenary 
sharpshooters," who take you captive by the spell of their 
wisdom, and will not let you go again without a ransom. 2 
He proceeds, accordingly, to restate the theory of Prota- 
goras in a form which secures it against cheap criticism of 
this kind. 

1 It is probable, indeed, that this is only Aristotle's inference from the 
Cratylus and the Theaetetus, but it is a fair inference. 

2 The reference to the Megarics is unmistakable here. The rift within 
the Sokratic school is evidently widening. 


183. As restated by Sokrates, the doctrine of Prota- 
goras is as follows. However true it may be that the 
sensations of each individual are his and his only (?<W 
/cacrTGj), and that what is (if the word is to be used at all) 
is what appears to the individual and to him alone, Prota- 
goras never intended to deny the distinction between wise 
and unwise. He would say that the wise man is one who 
is able to change bad beliefs to good. Belief, or what 
appears to one man, differs from belief, or what appears to 
another, not as true from false (for what appears to the 
individual is, and is therefore true and the only truth), 
but as good from bad, healthy from diseased, and the wise 
man is he who by his words can make what is good appear, 
and therefore be> good for the state and the individual 

Let us examine this. We shall see the bearing of it 
best if we consider questions of expediency or the advan- 
tageous (TO w<j)e\Lju,ov). In such questions it will be ad- 
mitted that one man is a better adviser than another, even 
by those who maintain that such distinctions as right and 
wrong are only conventional, that is, that they have no 
independent reality by nature, but depend for their 
existence and duration on the opinion of the community. 
No one, in fact, would maintain, except as a mere form of 
words, that what a state thinks advantageous for it is 
therefore advantageous for it This will be still more 
obvious if we consider the whole " form " (e?<Jo?) to which 
the advantageous must be referred. The general charac- 
teristic of it is that it has to do with the future. Now we 
may say that the present sensation of the individual is the 
only test (icpiTqptov) by which we can judge what is, but it 
will not be maintained that it is also the test of what is to 
be. With regard to that, the belief of the professional or 
the specialist always carries more weight than that of the 
layman. Where the future is concerned, it is not every- 
one, but the man who is wiser than others, who will be 
the " measure," and Protagoras himself admits this ; for 
he holds the wise man to be the man who can replace 


worse by better beliefs with regard to these very things. 
We see, then, that when we state the doctrine of Prota- 
goras sympathetically, it at once takes us beyond sensation- 
alism. It is no longer true, even according to him, that 
what appears to me is to me, and what appears to you is 
to you. This is specially noted (179 b) as the argument 
which is most fatal to the doctrine of Protagoras, though 
there is another which also disproves it. Protagoras must 
admit that the beliefs of other people are valid for them, 
and most other people do not believe the theory of Prota- 
goras to be true. Therefore it is not true for them. 1 

184. This piece of reasoning is interrupted by a 
magnificent digression on the philosophic life, conceived as 
it was in the Gorgias and the Phaedo. It is impossible to 
summarise a passage like this ; it must be read as it stands. 
Still, we are bound to ask ourselves why it is inserted here. 
It comes in the middle of a discussion intended to show 
that the wise man is the best judge of what is advantageous 
for the community, and yet it describes in glowing colours 
the, aloofness of the philosopher from practical concerns 
of every kind. The world is of necessity evil, and the 
philosopher will strive to escape with all speed from it to 
a better. The only way to do this is to become likened 
unto God, so far as that may be, and this likeness is to be 
attained by the cultivation of holiness and wisdom, and 
especially of geometry and astronomy. That is just the 
doctrine Plato consistently attributes to Sokrates, but it 
can hardly be an adequate representation of his own atti- 
tude to life at the time he wrote the Theaetetus. He was 
shortly to become involved in politics of a decidedly prac- 
tical nature, as we shall see, and the Academy was as much 
a school for statesmen and legislators as anything else. In 
the Timaeus Sokrates admits, as we have seen, that practical 
politics is something foreign to his interests, and we might 
therefore say that the present passage is inserted to keep 

1 This is the argument which came to be known as the TrepirpoTrrj or 
"turning the tables." It was also used against Protagoras by Demo 
kritos (Sext. Emp. vii 389). 


the picture of him true to life, at a time when Plato was 
entering on a course his master would have shrunk from 
instinctively. I believe that to be true, but it is not the 
whole truth. I believe that Plato, though he had learnt the 
duty of philosophers to descend in turn into the Cave, 1 still 
felt that the life here described was in truth the highest. 
It is not uncommon for a man of action to feel intensely 
the superiority of the contemplative life ; and it is not 
unnatural for such a man, if he is also a great artist, to 
sing the praises of what has become for him an impossible 
ideal, though he may recognise it in his inmost heart as 
saving truth. In the " digression " of the Theaetetus I 
think we may see Plato's reluctant farewell to the theoretic 
life. At any rate, he tells us himself that it is a digression 
unconnected with the main theme of the dialogue, and he 
must have had some motive for inserting it. 

185. We must now examine the claims of the theory 
of universal motion to give an account of knowledge. We 
must not forget that Melissos and Parmenides have asserted 
an exactly opposite theory, namely, that all is one and at 
rest in itself, having no space to move in. We stand, 
then, in a cross-fire between two hostile camps. Let us 
attack a the streamers" (ol peovres) first. We shall see 
that, on their theory, knowledge is impossible (179 d 
181 b). 

When we say " everything moves," what do we mean by 
" moves" ? There are two forms (etSy) of motion : (i) motion 
from place to place (<popa) ; (2) motion from state to state 
(oXAo/oxn?). In other words, motion is either locomotion or 
alteration ; and, if motion is universal, it must include both. 
Since, then, everything not only moves its place, but also 
alters its state, we cannot ascribe any quality to what moves ; 
for what we call qualities ('Troior^re?) are nothing but per- 
petual processes going on between what acts and what is acted 
upon, and accordingly, in the very moment of being named, 
the quality is gone. Similarly, as we may not speak of sen- 
sible qualities, so we may not speak of sensations ; for each 
sensation is in process, and cannot be called sight, hearing, or 

5200 : 


the like, any more than not-sight, not-hearing, and the like. 
And, if we cannot speak of sensation, we cannot speak of 
knowledge, which we identified with sensation, and the 
answer of Theaitetos was no answer, and the attempt to 
prove it by the theory of universal motion has only resulted 
in proving that all answers are equally right. In fact, we are 
not entitled to distinguish one answer from another; for such 
words as " thus " and " not thus " imply fixity, not motion 
(i 8 1 b 183 b). 

Sokrates declines to examine the " partisans of the 
Whole" (pi rou o\ov (TTao-iSrat), 1 Melissos and Parmen- 
ides, for the present ; we must come back to the original 
answer of Theaitetos. 

1 86. In ordinary language we speak of "seeing with 
the eyes," " hearing with the ears," and so on, but strictly 
we ought to say, not that the eyes are that with which we 
see (& optofjiev), but that they are the instruments (opyava) 
through which (<V Si/), or by means of which, we see. For 
we cannot suppose ourselves to be like so many Wooden 
Horses, each with a number of sensations sitting inside ; 
we must suppose that there is some one constituent 
element (ef<W) in us call it soul or what not in which 
all these sensations converge, and to which they serve as 
instruments when we are sensible of objects. This dis- 
tinction between the one identical element and the instru- 
ments employed by it may be made clear as follows. The 
instruments through which we are sensible of hot, hard, 
light, sweet things are various parts of the body. Each 
of these instruments has a specific power (Swajuu?), and 
that which one can do another cannot ; we cannot be 
sensible of sound by means of sight, nor of colour by 
means of hearing. If, then, we have a thought of any- 
thing which is common both to sound and colour, this 
must be due to some other instrument than seeing or 
hearing, and it is certain that we do have thoughts of 
things which are common to the objects of different senses. 
Let us see what these are (184 b 185 a). 

1 CL E. Gr. ?/5. 2 p. 140,0. i. 


To begin with, we have such thoughts as " colour and 
sound are," " each is other than the other and the same as 
itself," " both are two," cc each is one," " they are like or 
unlike one another/' and so on. What, then, is the power 
and what is the instrument through which it acts, by which 
we are enabled to find this common element to which we 
give such names as being and not-being (ova-la KOL TO w 
etvai), likeness and unlikeness (pnoton KOI^ avoy.oioTw\ 
sameness and otherness (TO TOUTOV re KOI TO Odrepov), unity 
and number (TO ev KOL TOV 3X\ov apidpov), odd and even 
(-repiTTov KOI 5/mov), fair and foul (jtaXoV KCU aurxpov), good 
and bad (aya6ov KCU KOKOV) ? Not one of these common 
properties (KOLVO) has any specific instrument by which it is 
apprehended, as was the case with such properties as sweet- 
ness, hardness, and so forth ; it seems rather that in those 
cases the soul is its own instrument (airr^ oY avrijs eTnovcoTra), 
and acts by itself (xad 9 CLVTVV). 

The simple sensation, then, of the sensible qualities of 
things takes place through the affections of the body (TO. 
TOV crco^crro? vad^fJLara) ; such sensation begins with birth 
and is common to man and beasts. On the other hand, 
the apprehension of the common qualities of things implies 
comparison and reflexion (TO ava\oylQ(r6ai 9 <jvXAoy*o7*6V, 
<rvfjL/3d\\eiv), whether of the most common property, that 
of being, or of those of sameness and difference and the 
rest, or of those of fair and foul, good and bad, the investi- 
gation of which last implies comparison in a pre-eminent 
degree in the bringing of past and present into relation 
with future, which requires time and effort and education 
(185 a i86c). 

It is at this point that we should expect Sokrates the 
Sokrates we have learnt to know from the Phaedo and the 
Republic to introduce the doctrine of incorporeal and 
intelligible forms; but nothing whatever is said about 
them either here or in any other part of the dialogue. 
Instead, we have the beginnings of a theory of what were 
afterwards called Categories, and these are regarded as 
certain common predicates which the soul apprehends 


without the instrumentality of sense, and by means of 
which it organises the manifold of sense. It is also to be 
observed that these common predicates apprehended by 
the soul alone include not only categories of reality 
(ova-la) 9 but categories of value (&<j>e\la). The practical 
is becoming more prominent than it was in the earlier 

187. Now, if there are predicates of this kind which 
are common to the sensations of all the organs of sense, 
and are apprehended by a purely mental activity, it follows 
that we cannot identify knowledge with sensation. The 
apprehension of being is essential to knowledge. Being 
and truth cannot be apprehended in the affections of the 
body, but only in the soul's reflexion about them. We 
must, therefore, look for knowledge under the name 
which describes the proper activity of the soul when 
it is concerned with what is. That name is judgement 
(TO $o%Aleiv). Is that to be identified with knowledge ? 
(186 c 187 a). 

The definition of judgement is not given till later, but 
it will be convenient to state it here. Thought (TO 
Siavoeiordat) is the discourse (StdXoyo?) that the soul holds 
alone with itself. When it has come to a determination, 
whether slowly or by a swift dart at a conclusion, and is at 
last at one and no longer at variance with itself, we call 
this its judgement ($6%a). Here we have a very remarkable 
change in terminology. In the Republic the word (&>), 
which is now used to signify the completed result of 
thought (Sidvoia*), means something lower than thought, 
and covers "imagination" (eiKaa-ia) and belief (TT/CTT/?). 
Plato is preparing to attack the problem of predication 
in his own way, and he wants a word for "judgement," 
and this seems the moat natural to take. We must 
understand the term here in the sense in which it is 
defined, and not in that which it bears in earlier dialogues. 
It is the characteristically Platonic as distinct from the 
Sokratic use of the word. It recurs in the later dialogues, 
and in certain Academic passages of Aristotle. We have 


to ask, then, whether knowledge is to be found within this 
activity of the soul. Does simple judgement contain in 
itself the guarantee of truth ? 

1 88. The second section of the Theaetetus is accord- 
ingly devoted to showing that no representation of the 
independent (aMi KO.& av-njv) action of the soul can be 
made to explain the undoubted fact of the distinction 
between true and false judgement. It is shown that 
thought alone is as incapable of yielding knowledge as 
sensation alone, nor is it clear how any combination of 
sensation and thought can yield knowledge. 

In the first place, we can only say that true judgement 
(a\r}6w 5oa) is knowledge. True judgement or thought 
is to judge something to be what it is ; false judgement or 
thought is to judge something to be other than it is. But 
this at once raises a difficulty. How can thought as such 
be other than true ? How can there be a false judgement 
at all ? So long as we confine ourselves to the independent 
activity of soul, it would seem that false judgement is as 
impossible as we have seen false sensation to be. Three 
possible accounts of it are examined, and are all found to 
be equally unsatisfactory. They either imply that it is 
possible to know and not to know the same thing at the 
same time, or that we can judge without judging any thing, 
or that it is possible to judge one thought to be another* 
To identify knowledge with the work of the mind is, 
therefore, open to the same objections as its identification 
with sensation. All judgements will be equally true, and 
the distinction between knowledge and ignorance, wisdom 
and unwisdom, will disappear. Thought, in fact, can be 
attacked with precisely the same weapons as sensation 
(187 b 190 e). 

189. It might seem more hopeful to regard true 
judgement as the reference of an impression of sense to 
the right or corresponding mental counterpart. We might 
suppose that memory is like a waxen tablet in the soul on 
which images are impressed. It is impossible that two 
impressions on this tablet should be confused, or that a 


sensation which makes an impression on it should be con- 
fused with another simultaneous sensation. It is, how- 
ever, possible that there should be error in the reference 
of a sensation to the memory-image left by a former 
sensation, if that image was not sharply impressed or if it 
has been worn out, That would be false judgement. 
This, however, is still unsatisfactory ; for it would restrict 
true judgement, and therefore knowledge, to judgements 
about actually present sensations. It would not explain, 
for instance, how some people can judge that 5 + 7 = 12, 
and others that 5 + 7 = 11, where there is no present 
sensation of such a number of objects. To explain this, 
we should have to make a distinction between having and 

possessing knowledge (et$ CTTLO-T^JUL^ and KTVO-LS 7ri(TTqM<i), 
of which the latter may exist without the former, just as 
we may possess a coat without actually having it on. Let 
us compare the mind to a dovecot in which we have shut 
up a number of birds that we have caught. We possess 
these birds, indeed, but we cannot be said to have them 
till we have caught them again. Now we may catch the 
wrong bird, and in the same way we may catch the wrong 
piece of knowledge, and that will be false judgement. 

Even that, however, is unsatisfactory, unless we suppose 
there are ignorances flying about in our mental dovecot 
also. But that will not do either ; for, when we have 
caught our bird, it is a bird in the hand and we know 
what it is. We are not any nearer an explanation of false 
judgement than we were before (191 b 200 d). 

Finally, it is certain that there may be true judgement 
without knowledge. The pleaders in the law courts 
operate by means of persuasion and not by means of 
instruction, and yet the jury may be led by them to form 
a true j udgement. This suggests to Theaitetos a definition 
which he has heard of knowledge, namely, that it is true 
judgement accompanied by a rational account of itself 
(a\^6rjg Soa fjLGra \6yov). Sokrates identifies this definition 
of knowledge with an elaborate theory he has heard " in a 
dream." There are some persons who maintain that the 


real is unknowable. Our sensations are produced by 
simple elements (o-ro^eFa) which are unknowable just 
because they are simple. They can only be named and 
cannot be defined, nor can we predicate anything of them, 
not even "being" or "this." Such properties as these 
are common to all sorts of things and cannot be regarded 
as properties of the simple reals. These can, however, be 
apprehended by sense, and we can give them names 
(oi/o/xara). They can also combine with one another just 
as letters (crro^^a) can form a syllable (<nAAa/3??). If we 
combine their names, we get a statement or proposition 
(Xo^o?), and that makes their combinations knowable 
(201 a 203 b). 

190. The "dream" of Socrates reminds us of the 
"mystery" of Protagoras, and we feel that they are both 
devices for going beyond historical verisimilitude. There 
is also the same difficulty about the authorship of this 
theory, as there is about that of the sensationalist theory 
described in the early part of the dialogue. In the first 
place, it must be observed that it is a thoroughly idealist 
theory in the modern sense of that word. The simple 
reals are themselves unknowable, and all our knowledge is 
the work of the mind. In this respect it is the exact 
counterpart of the earlier sensationalist theory. Thought 
is everything here as sensation was everything there. 
Now there can be no doubt that the definition of know- 
ledge as true judgement accompanied by a rational account 
of itself or ground (/uera Xoyov) belongs to the Sokratic 
school. It is the definition adopted by Diotima in the 
Symposium (202 a), and it is also taught in the Meno 
(97 e sq.}. It is more difficult to say where the elaboration 
of it we find here comes from. Aristotle appears to 
allude to it in a passage of the Metaphysics, in the course 
of which he makes a remark about the view of Antisthenes 
" and such uncultivated people " that it is impossible to 
define the " What is it ? ", because a definition would be a 
"long enumeration" (jmcucpos Xoyoy), and on the strength 
of this the whole theory has been attributed to Antisthenes. 


But all Aristotle says is that the theory in question appears 
to give plausibility to the view of Antisthenes, and, what- 
ever we may think of it, it is not a theory likely to have 
been set up by " uncultivated persons." l Antisthenes 
denied the possibility of predication, whereas, according to 
this theory, knowledge consists of nothing else. Nor is 
there any reason why Sokrates should cc dream " of 
Antisthenes. The suggestion made long ago by Lewis 
Campbell that the theory is that of " some Pythagorean " 
is much more plausible. 2 The terminology of letters 
(crroixeia) and syllables (cnAAa/3a/) is characteristic of the 
Pythagoreans, and we can see quite well how these 
Pythagoreans who refused to adopt the Sokratic doctrine 
of the participation of sensible things in the forms might 
find themselves driven to some such theory as this. In 
any case, the importance of the discussion is missed 
altogether if it is not clearly understood that the doctrine 
discussed is the exact opposite of the sensationalism 
Protagoras is said to have revealed "in a mystery/' and 
that it is rejected as equally unsatisfactory. 

191, For, when we come to examine it, we find that 
this theory leads to very great difficulties. How are we 
to conceive the relation between the prime elements and 
the complexes which are the objects of knowledge ? 
Either the syllable is only the sum of the letters, in 
which case it is impossible to see how it should be more 
knowable than they are, or it is an indivisible unity, in 
which case it cannot be known either, since that would 
imply the separate apprehension of its parts. 

Further, we must ask precisely what we mean by an 
" account " (Ao'-yor) in this connexion. Obviously we do not 

1 Met, B, 3. 1043 b, 5 sqq* Antisthenes is not mentioned till b, 24, and 
the passing manner in which he is alluded to seems to me to exclude 
the idea that Aristotle was thinking of him at all when he began the 

2 Introduction to the Theaetetus, p. xxxix. The theory would 
harmonise well enough with what we are told of the doctrine of 
Ekphantos of Syracuse, 


mean merely the expression of a judgement in articulate 
language. Nor can we mean a simple enumeration of the 
elements which make up a thing. Rather, we must mean 
a statement of the thing's differentia (Siatpoporys), that 
which marks it off from all other things. If, however, 
we mean by this that we have merely a judgement (5oa) 
as to the differentia, that brings us no further forward ; 
while, if we mean that we have knowledge of the differentia, 
our definition will be circular. "True judgement with a 
knowledge of the differentia " is not a definition of know- 

The conclusion of the Theaetetus, then, is that knowledge 
can neither be sensation nor the work of the mind. Sensa- 
tion is merely a resultant of motion, and gives us no 
reality outside itself. Thought alone merely yields com- 
binations of names. Nor have we been able to show, 
except by clumsy images, how knowledge can be due to 
any combination of sensation and thought. On the other 
hand, we have incidentally made several discoveries as to 
the nature of knowledge. We have found, in the first 
place, that it implies certain " common " or generic predi- 
cates, and, secondly, that to know a thing we must know 
its differentia. A mere apprehension of its common pro- 
perties would not be an apprehension of it at all. The 
next dialogue we have to consider really deals with the 
same difficulties, though from another point of view. 

The Parmenides. 

192. The Parmenides is a criticism of the doctrine of 
forms as stated in the Phaedo and Republic, and the selec- 
tion of Parmenides as the chief speaker points to the con- 
clusion that the objections to the theory of participation 
contained in the first part of the dialogue are of Eleatic 
origin. We know from the Theaetetus that Plato was 
busy with Eukleides about this time. Besides that, we 
have a remarkable piece of external evidence to the same 
effect. The most telling argument against participation 


is that known as " the third man/' which we shall come to 
presently. We have unimpeachable evidence that this 
argument was introduced in some work or other by the 
" Sophist " Polyxenos. 1 He had been a pupil of the 
" Sophist " Bryson, who had been an associate of Sokrates 
along with Eukleides, and with him had founded the 
" Eristic " of Megara. He also stood in close relations 
of some kind with the Academy. 2 Now the detractors of 
Plato asserted that he plagiarised the lectures (Siarpi/3cu^ 3 
of Bryson, and that is most easily explained if we assume 
that Bryson was the original author of this argument. 

But, if these arguments are Eleatic in origin, it follows 
that they are not directed against the reality of the intel- 
ligible, but against that of the sensible. It would have 
been absurd to make Parmenides the mouthpiece of an 
attack upon the One, and all we know of the Megaric 
doctrine goes to show that it denied all reality to the 
world of sense. The arguments of the Parmenides are 
not directed, then, against the doctrine of forms as such, 
but against the Sokratic theory that sensible things come 
into being and cease to be by partaking or ceasing to par- 
take in the forms. An argument like the " third man " is 
clearly double-edged. It may be used to show the impos- 
sibility of an avrodvdowTros, but it will serve equally to 
demonstrate the unreality of particular men. Plato was, 
of course, far too interested in the world of experience to 
accept the acosmism of Eukleides, but he was clearly 
impressed by the force of the arguments against <c partici- 
pation " as an account of the relation between the sensible 

1 Alexander on Ar. Met. 990 b, 17. He quotes Polyxenos from 
Phanias of Eresos, a disciple of Aristotle and friend of Theophrastos. 
See Batimker in Rhein. Mus* xxxiv. pp. 64 sqq. The word ctcrayetv used 
by Phanias does not necessarily imply that Polyxenos invented the 
argument. Cp. etcrayetv, " to bring on the stage/' 

2 This appears from the comic poet Ephippos, fr. 14 Kock. It is not 
clear whether Bryson was a member of the Academy, but he may have 
been. It makes no difference. What is important is that he was an 
associate of Sokrates. 

8 Theopompos, af. Athen. 509 c. 


and the intelligible. His own account of that is not, 
however, given in the Parmenides. 

193. The subject of the dialogue is introduced as 
follows. One of Zeno's arguments against the opponents 
of Eleaticism was that " if things are a many, they musi 
be both like and unlike." The precise meaning of this 
does not concern us here ; what we have to deal with is 
the solution of the difficulty proposed by Sokrates, who is 
not an old man, as in the Theaetetus, but " extremely 
young" (127 c). He asks Zeno whether he does not 
believe in "forms" which are "apart from" the things 
of sense, but in which these things " participate." If that 
is the truth, there is no reason why sensible things should 
not participate at once in the form of likeness and in the 
form of unlikeness. A man, for instance, is both many 
and one ; he has many parts, but he is one man among 
others. Why should not a sensible thing be at once like 
one thing and unlike another, thus partaking in both 
forms ? To show that stones, sticks, and the like are 
both many and one is not to show that One is many 
or Many is one. What would be surprising would be 
if a man should set up separate forms such as Likeness 
and Unlikeness, One and Many, Motion and Rest (i.e. the 
common predicates (KOIVO) of the Theaetetu$\ and should 
then show that these can mingle with and be separated 
again from each other. It would be still more surprising 
if he could show that the same contradictions which have 
been shown to exist in the things of sense were also to be 
found in forms apprehended by thought (129 a 130 a). 

The theory here stated by Sokrates is precisely that of 
the Phaedo, where we are told that Simmias may be greater 
than Sokrates and smaller than Phaidon, though Greatness 
and Smallness exclude one another (102 b). It is to be 
noted, however, that, even in the Phaedo, a doubt is 
expressed as to the adequacy of the term " participation," 
for the relation between a subject and its predicates (100 d). 
If the Phaedo is in substance historical, it will follow that 
the Sokrates of the Parmenides is just Sokrates himself 


before he had begun to feel these doubts. That Plato 
should have meant his own earlier self will only be 
credible to those who can believe that in the Phaedo he 
made use of Sokrates as a mask for his own personality, 
while the view that by Sokrates here he meant some 
callow Academic who held his own theory in a crude 
form should be credible to no one. We might be reluc- 
tantly convinced that Plato used Sokrates as a disguise 
for himself; but it would surely have been impious to 
represent his own immature disciples under the revered 
name of his master. The fact that it has to make 
assumptions of that kind ought to be fatal to this line 
of interpretation. 

1 94. Parmenides, who has evidently heard of " forms" 
before (130 a), and who is delighted by the philosophic apti- 
tude of Sokrates, as shown by his theory of " participation," 
begins by asking him whether, in addition to the mathe- 
matical forms, which have been mentioned so far, he also 
believes in forms of the Just, the Beautiful and the Good, 
and, as might have been expected from the Phaedo, Sokrates 
at once assents. The next question is whether he believes 
in forms of Man, Fire, and Water. Sokrates confesses 
that he is in a difficulty about these. We have seen what 
this means ( 73). As to things like mud, hair, and dirt, 
though he has sometimes been troubled by the thought 
that they must have forms too, he had finally renounced 
the idea. That, says Parmenides, is because Sokrates is 
still young, and philosophy has not yet laid hold of him 
completely as it will do some day. Then he will despise 
none of these things ; at present he is too much influenced 
by popular opinion (1306). 

In the mouth of Parmenides this remark must be ironical. 
He must mean that, if such things as hair, mud, and dirt, 
are in any sense real, they are quite as much entitled to 
have "forms" as the objects of mathematics. From Plato's 
point of view, on the other hand, the passage has probably 
another bearing. The doctrine of forms, as hitherto stated, 
is only plausible because it is confined within certain limits. 


It is adequate in mathematics, where it originated, because 
in that region even the particulars are objects of thought 
and not of sense. In morals and aesthetics it is almost as 
satisfactory ; for actions in their moral aspect are not really 
objects of sense, and beauty is a direct revelation of the 
form. On the other hand, it is a serious weakness in the 
doctrine that it can only be applied with difficulty in 
physics and biology, and that it breaks down altogether 
when we come to things common and unclean. If, now, 
we remember the way in which Plato insists in the 
Theaetetus on the distinction between the "common" 
predicates (KOLVO) which the soul apprehends by itself, 
and the objects of the several senses, we shall be inclined 
to think that he is preparing the way for a restriction of 
the doctrine to the former, while suggesting at the same 
time that this very restriction may so modify the doctrine 
that it will enable us to understand the whole world of 
experience, even in its humblest manifestations. There is 
no inconsistency in the restriction of the doctrine to purely 
intellectual categories, and the extension of the operation 
of these categories to the whole of the sensible world. 
Nor is any weight to be attached to the fact that in the 
Timaeus we have forms of Fire and the other elements ; 
for there the speaker is a Pythagorean, and we have seen 
reason to believe that it was just in the construction of 
the elements that the later Pythagoreans made most use of 
the forms. 

1 95. Leaving this question for the present, Parmenides 
goes on to discuss the difficulties involved in the specially 
Sokratic conception that the many sensibles "partake in" 
the one form, or that the one form is "present to" or 
"in" the many sensibles. 

In the first place, these sensibles must either all contain 
the whole of the form or each of them only a part of it. 
In the first case, the whole form will be present in each 
particular thing, which means that it will be in more pla.ces 
than one, and so will be separate from itself and divided. 
Sokrates suggests that it may be like the day, which is 


present in many places and yet one, but Parmenides will 
not accept this comparison. If a number of people are 
covered by the same sailcloth, each one of them is covered 
only by a part of it. We come, then, to the other alter- 
native, that the forms are divisible, and that what partakes 
in a form contains only a part of it ; or, in other words, 
that only a part of the form is present in each of the many 
sensibles. In that case, however, the forms will not serve 
to explain anything. A part of the form of magnitude, 
if there could be such a thing, would be less than the 
whole, and a thing could not become great by participating 
in it, and many other absurd consequences would follow. 1 

Further, the very grounds on which Sokrates bases the 
doctrine of the one form in which innumerable sensible 
things partake would really compel him to assume also the 
existence of equally innumerable forms. If we require a 
form to explain the participation of particular things in a 
common predicate, we also require a form to explain the 
participation of the form itself and the particular things in 
a common predicate, and so on ad infinitum (132, a). 

Sokrates hereupon suggests that perhaps the forms are 
really thoughts (i/o^ara), and that they may only exist in 
souls, to which Parmenides replies that a thought must be 
a thought of something real, and further that, if the forms 
are thoughts, the things that partake in them must be 
thoughts too. It would also follow either that all things 
think or that there are unthought thoughts. 2 

The next suggestion made by Sokrates is that the forms 
may be "patterns" (TrapaSelyjmaTa), and that the true 
account of the participation of sensible things in them may 
be that they are "likenesses" ( a) of them. 3 But, 

1 For the details of these I must refer to Professor Taylor's article in 
Mind (N.S.), vol. xii. No. 45, 

2 The last point is somewhat obscure, but it does not affect the main 
argument. Observe how clearly Conceptualism is formulated, and how 
deliberately it is rejected. 

8 According to Aristotle this was the Pythagorean view (Met. A. 6). 
We can, therefore, draw no inference from its prominence in the Timaeus, 


says Parmenides, if the things are like the forms, the forms 
will be like the things, and we shall require another pattern 
which both resemble to explain their likeness. We are 
confronted once more by an infinite regress. 

But there are far more serious difficulties than these. It 
would be very hard to refute anyone who said that these 
forms, if they are such as we describe them, are unknow- 
able. We have said that they are "alone by themselves" 
and not in our world (ev ay/wi/), and therefore, as they are 
relative by nature, they can only be relative to one another. 
On the other hand, their " likenesses " in our world can only 
be relative to one another and not to the forms. A man 
is not the master or slave of " mastership itself" or of 
" slavery itself/' but of another man ; while, on the other 
hand, "mastership itself" is relative to "slavery itself," 
and not to a particular slave. In the same way " know- 
ledge itself" is relative to " truth itself," but our knowledge 
is relative to the truth in our world. But, if that is so, 
the forms must be entirely unknown. If we try to avoid 
this by saying that God has " knowledge itself," and there- 
fore knows the forms, the result is still worse. It will 
follow that God cannot know us or anything that we 
know; for the knowledge he has is not relative to the 
truth of our world. Nor can he be our Master; for 
" mastership itself" is not relative to us (134 d-e). 

196. This section is based on the argument of the 
"third man," which has already ( 195) been used to throw 
doubt upon the theory of participation. It will be well 
to give it here in the form in which Phanias of Eresos 
quoted it from Polyxenos. 1 " If a man is a man in virtue 
of participation or partaking in 2 the form or the at/ro- 

wliere the speaker is a Pythagorean, least of all the inference that Plato 
himself adopted this view in later life. 

1 See above, p. 254, n. I. 

2 It is important to notice that Polyxenos uses for "participation" 
two terms (^roxrj, perowia), which are never used by Plato. That 
goes to show that the argument was not specially directed against Plato's 
statement of the theory. 


s, there must be a man who will have his being 
relatively to the form. Now this is not the avrodvOpooTros, 
who is the form, nor the particular man who is so in virtue 
of participation in the form. It remains, then, that there 
must be a third man as well who has his being relative to 
the form." 1 I understand this to mean that, as it is im- 
possible for the particular sensible man to stand in any 
relation to the form, and, as the form cannot be related 
simply to itself, the theory of participation explains nothing. 
The only * c man " who could participate in the form of 
Man would be a third man in the intelligible and not in 
the sensible world, and it is quite superfluous to assume 
anything of the sort. It will be observed that, as has been 
suggested above, this argument is directed against the 
reality of the sensible and not of the intelligible. It is first 
and foremost an argument against the theory of participa- 
tion, and it is only an argument against the doctrine of 
forms in so far as that implies many particular forms of 
man, etc., instead of a single absolute One. That explains 
further how it is that, while Aristotle uses the argument 
against the doctrine of forms, he also thinks it necessary 
to refute it. 2 It was intended to support a position with 
which he had still less sympathy. 

197. It almost seems as if we should be driven to the 
conclusion that the forms are unknowable, and that would 
be the end of all philosophic discussion. It would destroy 
dialectic (ryv rov SiaXeyea-Ocu Svva/juv). It Is hinted, indeed, 
that a solution may be found (135 a), but this is not 
followed up for the present. Instead oif that, Parmenides, 
who could hardly be expected to undertake the task of 
justifying the world of experience, proposes to dismiss 
that from consideration altogether, and to consider the 
difficulties that arise in the world of forms itself. The 
argument is still on Megaric ground ; for we know that 

1 1 have adopted the transposition of Baumker (3(hein. Mus. xxxiv. 
P- 75)- 

*SofJLl m 1780, 


Eukleides rejected the multitude of forms and reduced 
them all to the One. 

At the beginning of the dialogue (1290^.) Sokrates 
had declared himself unable to understand how the forms 
themselves could enter into combinations with one another, 
and still more how a form can be both one and many, like 
and unlike, at rest and in motion. It is easy enough, he 
repeats here (135 e), to see how sensible things can have 
different predicates ; the real difficulty arises when we 
apply this to the forms. The way to deal with a problem 
of this kind, says Parmenides, is the method of hypothesis, 
and that both in its positive and negative application. We 
must trace out all the consequences (o-v/uipalvovTa) of the 
hypothesis that it is and also of the hypothesis that it is not. 
For instance, if we take the hypothesis Zeno examined, 
"If things are a many . . .", we should go on next to 
the consequences of the hypothesis "If things are not a 
many . . .", and in both cases we should ask what are the 
consequences, not only to the subject of the hypothesis 
itself, but also to the rest, and in each case we should 
consider the consequences to the subject of the hypothesis 
and to the rest both in themselves and in relation to one 
another. The same method must be followed in the case 
of all the forms, such as likeness and unlikeness, rest and 
motion, coming into being and ceasing to be, being and 
not being, and so forth (or, in other words, the "common" 
predicates of the Theaetetus). 

198. Parmenides naturally takes his own doctrine of the 
One as the hypothesis to be examined. Plato has his own 
reasons for this, as we shall see, but there is no ground for 
thinking that either Parmenides or Sokrates is supposed 
to be conscious of them. Parmenides is not represented as 
accepting the consequences of his argument he could not 
do that without destroying his own system and he ex- 
pressly declares that the result of his examination of the 
first hypothesis is impossible (142 a). Sokrates is reduced 
to silence, but we cannot suppose him to be convinced. 
The whole thing is treated as a mental gymnastic 


a "laborious game," valuable chiefly for the training it gives 
in method. Plato means more than that, however, and he 
gives us the hint in the dialogue itself. We must remem- 
ber that the discussion is about forms alone, and we are 
expressly warned against the idea that "the rest" of which 
he speaks are the things of sense (135 e). They are just 
the other forms. Now Sokrates had said (129 d sqq.} that 
he would be very much astonished if anyone could show 
that the forms were capable of combination with one 
another. That form of separation (xwpLa-fjLoi) had been 
clearly taught in the Phaedo, for instance. Sensible things 
could participate in the forms, but the forms excluded one 
another. He would be still more astonished, he adds, if 
anyone could show that there was the same sort of con- 
fusion and uncertainty in the forms as there is in the 
sensible things which participate in them, and that is exactly 
what Parmenides does show. If you take such forms as One 
and Being abstractly (x^)> ^ e y a ^ once partake of and 
begin to pass into one another and all the other forms, 
including even their opposites. They are just as bad as 
water, which is cold to one hand and hot to the other, or 
any other of the sensible things which we have seen to be 
in continual flux. In fact, Parmenides proves that, if we 
take the intelligible world by itself, it is quite as unsatis- 
factory as the sensible, and by taking the One as his 
example, he really refutes the Megaric doctrine, and that 
with the weapon of the Megarics themselves. It adds to 
the humour of the situation that this refutation is ruthlessly 
carried out by the revered Parmenides, and it is even possible 
that we are to regard the description of his own work given 
by Zeno in the introduction as a hint of the light in which 
Plato wishes us to look at the second part of our dialogue. 
Zeno says : 

My work makes no sort of pretence to have been written 
with the object you mention (i.e. to prove the doctrine of 
Parmenides in another way). ... It argues against those 
who maintain a multitude, and gives them back as good or 
better than they gave, by trying to show that their hypothesis 


will have even more absurd consequences than his, if it is 
thoroughly discussed (128 c-d). 

Just so we may say that Plato has no idea of proving the 
hypothesis of his master, Sokrates, but he does propose to 
show that the hypothesis of the Megarics has even more 
absurd consequences than his if it is adequately followed 

199. It is from this point of view we must judge what 
strikes a modern reader as the arid and repellent form of 
the discussion with its occasional suggestion of sophistry. 
It is a display of the dialectical method introduced by Zeno 
and assiduously cultivated by his successors at Megara. 
Now Plato's dramatic power is by no means extinguished 
yet, and whatever impression it makes upon us, we may 
be sure that his contemporaries would keenly appreciate 
the virtuosity with which he plays on this alien instrument. 
It should be added that, so far as the arguments are 
sophistical and one or two of them must certainly have 
been known by Plato to be so that is probably quite 
deliberate. We shall see that he was coming to regard the 
disciples of Eukleides more and more as " eristics," just 
because, as we saw in the Theaetetus^ arguments confined 
to the objects of thought alone consist of judgements which 
are only combinations of names. There is, in fact, no 
dialogue where it is more important to remember the 
dramatic character of Plato's writing than this, and where 
it is more important to realise the contemporary situation. 
It seems to me quite possible that to Plato's circle the 
second part of the Parmenides seemed highly entertaining. 
Men who had laughed at the Euthydemus would find a 
subtler enjoyment here. I suspect, however, that Bryson 
and his friends were not pleased. In introducing Helikon 
some years later to Dionysios II. as a disciple of Eudoxos, 
Isokrates, and Bryson, he says, 1 " And what is rare on the 
top of all this, he is not unpleasant to deal with, and he 
does not strike me as malicious, but rather as easy-going 

1 Ep. xiii. 360 c. 


and simple. I say this with fear and trembling, seeing 
that I am expressing a judgement about a man, and man 
is not a bad animal, indeed, but a changeable one." We 
shall have occasion to note other traces of the growing 
estrangement of Plato from the Megarics. Let us now 
consider the hypotheses. 1 

200. There are properly speaking eight hypotheses to 
be examined, but there is a sort of corollary to the first 
and second, so that there appear to be nine. 

Hypothesis L If if is One y what will be the consequences 
for itself? (137 c). 

If it is One, it cannot be Many, and therefore it cannot 
have parts, and cannot be a whole (for that implies parts). 
Not having parts, it cannot have beginning, middle or end ; 
it has therefore no limits and is infinite. Further, it will 
have no figure ; for figure implies parts, Further, it will be 
nowhere ; for what is anywhere must either be contained in 
something else or in itself. It cannot be contained in anything 
else ; for it would then be in contact at different points with 
what contained it, and that implies parts. Nor can it be con- 
tained in itself; for then it would be both container and 
contained, and so two, not one. 

It cannot be in Motion or at Rest. If it suffered alteration 
(aAXo/om?), which is one form of motion, it would no longer 
be one. It cannot have spatial motion (0opa), which is the 
other form of motion, either motion of rotation (-Tre/H^opa), 
for that implies a centre or axis of rotation, and so figure and 
parts, or motion of translation, since it has no place. Further, 
it would have to be at once in the same place and not in the 
same place, which implies parts. Nor can it be at rest, since 
it is nowhere in space, neither in itself nor in anything else, 
and cannot therefore be where it is (ei/ rauro)). 

Nor can it be the Same as or Other than itself or anything 
else. It cannot be other than itself, for then it would not be 
one ; it cannot be the same as anything else, for then it would 
be the same as what is other than one j it cannot be other 
than anything else, for it is only the other that can be other ; 

1 1 have thought it right to analyse these somewhat fully as a guide to 
students of the Parmenides* From what has been said, it will be clear 
that the reader may omit them if he likes. 


it cannot be the same as itself, for if same were one, how 
could anything be the same as many ? 

It cannot be Like or Unlike itself or anything else, for the 
like is what has an identical property, and the only property 
of what is one is to be one. 

Nor can it be Equal or Unequal to itself or anything else. 
If it were equal, it would have the same measures, but it does 
not participate in the same. If it were unequal (greater or 
less), it would have as many parts as measures, and so would 
not be one. 

It cannot be older or younger than itself or anything else, 
or the same age, since all these imply inequality or equality. 
It cannot, therefore, be in time at all ; for what is in time is 
always becoming older than it is at a given moment, and 
therefore at the same time younger than it is, and also, since 
this becoming lasts no longer or shorter time than what 
becomes, it is always the same age as itself. 

Further, since it does not participate in time, it does not 
participate in Being; for it has not become and has not been, 
it will not become and will not be, it is not becoming and it 
is not. 

And, if it cannot be^ it cannot be one, and cannot be named, 
spoken of, judged of, known, or perceived by the senses. 

As this result seems impossible, let us put the hypothesis 
in another form. Let us consider One, not merely as 
one (TO eV ev ), but as being (TO ov ev). 

Hypothesis IL If One zV, what are the consequences for 
itself? (142 b, i 1550, 3). 

If One is, it partakes in Being (for is and one do not signify 
the same). Therefore One as being (eV ov) must be a whole 
of which one and being are parts. But, since each of these 
parts partakes in turn both of one and being, each can be 
further subdivided into two parts, and what always becomes 
two is not one but an infinite multitude. 

Again, if we take One by itself, it is other than being. 
But One is not Other, and Being is not Other, therefore 
Other is other than either. Any pair of these three must be 
called two or both, and each of two is necessarily one. If we 
add One to any of these pairs, we get three, and three is odd 
while two is even ; and two gives twice and three gives thrice, 
so that we have twice two and thrice three and twice three 
and thrice two. And so we may get any combination of odd 


and even numbers, and thus an infinite multitude, every part 
of which partakes in Being, so that Being is infinitely divided 
into parts. But each of these parts is one, so One is divided 
into as many parts as Being, and therefore not only One as 
being but One as one is an infinite multitude. 

One as being is a whole, and parts are only parts as parts 
of a whole, and the parts are contained in the whole. Now 
that which contains is a limit. But, if it is limited, it will 
have extremes, and, if it is a whole, it will have beginning, 
middle and end. But, as the middle is equally distant from 
the extremes, it will have figure, either rectilinear, or circular 
or mixed, and will be finite. 

Further, since all the parts which make up the whole are 
contained in the whole, it must be in itself; and, since the 
whole is not contained in the parts, it must, regarded as a 
whole, be in something else. Therefore it will be both at 
Rest and in Motion. 

Further, it will be the Same as itself and everything else, 
and Other than itself and everything else. It is other than 
itself because it is both in itself and in something else, and 
other than everything else, since these are not one. But it is 
also the same ; for otherness cannot be a property of anything. 
Therefore One and what is other than One, cannot be other 
because of otherness, nor can they be so in themselves. Nor 
can they stand in the relation of whole and parts ; for what 
is not One does not partake in number. Therefore they are 
the same. 

Consequently, it must be Like and Unlike itself and every- 
thing else, for One is other than everything else in the same 
way as everything else than One, and therefore they are alike 
in so far as they are other. On the other hand, they must be 
unlike in so far as they are the same ; for opposite antecedents 
must have opposite consequences. 

Further, it will be in contact with itself and with what is 
other than itself, since it is contained in something other. 
But, as contact always implies at least two, since the number 
of points of contact is always one less than that of the things 
in contact, it cannot be in contact either with itself or 
anything else. 

Further, it will be Equal and Unequal to itself and every- 
thing else. If it were smaller, Small would be in it, either as 
a whole or in a part of it. If it were in it as a whole, it 
would either pervade it completely, in which case it would be 
equal to it, or exceed it, in which case it would be greater. 


And the same contradiction arises if it is in a part of it. The 
same applies mutatis mutandis to the Great. Besides, Great 
and Small are relative to one another and not to One. 
Therefore One is equal to itself and to what is other than 
itself. But One is in itself, and therefore contains and is 
contained by itself, and is therefore greater and smaller than 
itself. And, since there is nothing besides One and what is 
other than One, and, since everything that is is in a place, 
what is other than One is in One, and One is therefore 
greater than what is other than One. But, for the same 
reason, One is in what is other than One, and therefore 
smaller than it. The same reasoning will apply to the parts 
as to the whole. 

Further, it will participate in time ; for it zV, and to be is 
just participation in being along with present time. But as 
time (of which the present is a part) is always advancing, One, 
as sharing in this advance, is always becoming older, and 
therefore at the same time younger, than itself. But it cannot 
advance from past to future without passing through the 
present ; and so, when it comes to the present, advance is 
arrested, so that the growing older and younger are already 
complete in the present. But the present lasts for the One 
as long as it is ; for it is always now whenever it is. There- 
fore the present lasts as long as time for the One, and its 
being older and younger coincides with its becoming older 
and younger. Further, since it is not and does not become 
for a longer time than it is and becomes, it is always the same 
age as itself. 

In the same way it is older than what is other than itself. 
What is other than One must be more than One, and being 
a multitude must partake in number, and One comes into 
existence before all other numbers. But it is also younger 
than what is other than One ; for it has beginning, middle, 
and end, and the beginning comes first into existence and the 
end Iast 3 and One only is when the end has come into 
existence. Therefore One only comes into existence after 
its parts. On the other hand, each part is itself one, and so 
One came into being simultaneously with the beginning and 
with every subsequent part, and must therefore be the same 
age as what is other than One. 

So much for its having become and being older and younger 
than what is other than One ; we have still to consider its 
becoming older and younger. On the one hand, it does not 
become either older or younger than what is other than One ; 


for, if the difference of two ages is given, the addition of equal 
to unequal times does not alter the (arithmetical) ratio between 
them. On the other hand, it does become older and younger ; 
for, if the difference of two ages is given, the addition of equal 
to unequal times does alter the (geometrical) ratio between 

Therefore One partakes of past, present, and future ; it 
was, it is, it will be 5 it has become, is becoming, and will 
become. It can be the object of knowledge, judgement, and 
sensation ; it can be named and spoken of. 


We have seen that One is (i) one and many and neither 
one or many, and (2) that it partakes in time. We must 
now consider how the second conclusion affects the first 
(155 e, 4 sqq.). 

If One is both one and many, and also partakes in time, it 
follows that it partakes in being at one time, viz. when it is 
one, and that it does not partake in being at another time, 
viz. when it is not one. To begin to partake in being is to 
come into being, to cease to partake in it is to perish ; there- 
fore One must come into being and cease to be (yeveaig KOL 
<j>6opd)* Therefore it must be compounded and decomposed 
again ; it must be assimilated and dissimilated again ; it must 
increase and decrease again and be equalised. 

Further, it must pass from motion to rest, and again from 
rest to motion. But how is that possible ? How can it stop 
when it is moving, or start moving when it is at rest ? The 
transition from rest to motion or from motion to rest cannot 
be either rest or motion, and there is no time at which a thing 
is neither at rest nor in motion. Therefore the transition 
must be out of time altogether ; it must be in that strange 
thing (TO CLTOTTOV TOVTO), the instantaneous (TO ^al<pvrj^) y 
which has position but not duration in time. It is the instan- 
taneous which makes all changes from one opposite to another 
possible, and it is in the instant of change that what changes 
has neither the one nor the other of its opposite qualities 
(155 e 157 b). 

Hypothesis III. If One is, what are the consequences for 
the others? (157 b, 6 159 b, i). 

The others are other than the One, but they will partake 
in it both as a whole and as parts. For, since they are others, 


they are a multitude, and this multitude must have parts or it 
would be one. Again, it must be a whole and a whole must 
be one. For, if a whole were not one but many, each part 
would be part of a many of which it itself was one. Then 
each would be a part of itself and of each of the others, which 
is absurd. Therefore they are a whole, that is a complete 
one made up of them all. Further, each part is also one since 
it is distinct from the others. Therefore both as a whole and 
as parts the others partake in One. 

Therefore they will be both finite and infinite. For, since 
they are more than one, they must be an infinite number ; 
for, if we cut off in thought the smallest imaginable portion 
of what is distinct from One, it will be more than One, and 
therefore an infinite multitude. On the other hand, at the 
moment when any part partakes in One, it has a limit both 
with the other parts and with the whole, and the whole has 
in the same way a limit with the parts. Therefore it is 

So too they will be both like and unlike each other and 
themselves. As being all finite and all infinite they are like ; 
while, as being both at once, they are unlike. And in the 
same way it would be easy to show that they are the same 
and other, at rest and in motion, etc., etc. 

Hypothesis IV. If it is One, what are the consequences 
for the others? (159 b, 2 160 b, 4)* 

The others will participate in the One neither as a whole 
nor as parts. For, since there is nothing which is at once 
other than one and other than the others (for One and the 
others are everything), One and the others cannot be con- 
tained in the same thing. Therefore they are quite apart. 
Further, since One as such has no parts, no part of it can be 
in the others. 

Further, since the others do not participate in One either 
as a whole or as parts, they are not a whole. Nor can they 
have multitude or number; for number consists of ones. 
Therefore they cannot have two properties, such as likeness 
and unlikeness, to One, nor even one property in themselves, 
such as Same, Other, Rest, Motion, etc. ; for that would imply 
participation in One. 

201. The result of our positive hypotheses, then, is 
this, One is everything and nothing both in itself and in 


relation to the others, and the same is true of the others. 
We now turn to the negative hypotheses. 

Hypothesis V. If One is not^ what are the consequences 
for itself? (160 b, 5 163 b, 6). 

If we can say that One is not, One must have a meaning, 
and therefore it must be knowable and there must be know- 
ledge of it. And, as it is other than everything else, it must 
have altereity (Irepoior^?). And it must partake in " this," 
" that," " anything," etc. ; for otherwise it could not be spoken 
of, nor could what is other than One be spoken of. There 
is nothing to hinder it partaking in many things, even if it is 
not. On the contrary, it must do so, if it is that One and 
can be named at all. 

Further, in so far as it is other, it must be unlike the others 
and like itself. 

Further, it must be unequal to the others ; for, if it were 
equal, it would be, and would be in so far like them. 

On the other hand, since Great and Small belong to the 
Unequal, and what possesses inequality must possess them ; 
and further, since the possession of Great and Small implies that 
of Equal as a necessary intermediate, it will possess all three. 

Further, it will participate in Being. For, if it is true that 
the One is not, then the One is a not-being. The very bond 
of its not being is that not-being is, just as the bond of what 
is is the not being of not-being. 

But, if it has both being and not-being, there must be a 
transition, that is, a movement from the one to the other, and 
this movement must imply alteration (aXAc/wer*?). 

On the other hand, One, so far as it is not, and therefore 
is in no place, cannot move from place to place, nor move in 
the same place round a centre. Nor can it alter without 
ceasing to be the One which is distinct from the others. 
Therefore it is immovable and unalterable. 

Further, it follows that, in so far as it is moved and altered, 
it comes into being and ceases to be ; in so far as it is unmoved 
and unaltered, that it neither comes into being nor ceases 
to be. 

Hypothesis VI. If there is no One, what are the con- 
sequences for itself? (163 b, 7 164 b, 4). 

If there is complete absence of being from One, it can 
neither partake nor cease to partake in Being. Therefore it 


can neither come into being nor cease to be ; it can neither 
be in motion nor at rest ; it cannot stand in any relation to 
what is, for that would be to partake in Being. Therefore it 
has neither greatness or smallness or likeness or unlikeness to 
itself or anything else. Neither is it in a place or in a time. 
Neither can there be knowledge, judgement or sensation of it ; 
it cannot be spoken of or named. 

Hypothesis VII. If One is not, what are the consequences 
for the others? (164 b, 5 165 e, i). 

Since they are others, they must have something that they 
are other than. They cannot be other than One ; for One 
is not. Therefore they must be other than themselves. 

Further, they must be so, not as ones, but as multitudes or 
masses, of which each can be broken into an innumerable 
number of similar parts, so that we can never reach a smallest 
and least part, and that what seemed small appears great com- 
pared with each one of the multitude of which it is the 

Further, we never come to a beginning, middle, or end, but 
always to something before the beginning or after the end or 
in the middle of the middle. 

The conclusion is that, if One is not, other things will 
appear both finite and infinite, one and many. 

Hypothesis VIII. If there is no One, what will be the 
consequences for the others? (165 e, 2 166 c, i). 

They will be neither one nor many ; for many implies ones. 
Nor have they even an appearance of one or many ; for they 
can have no communion with what is not, nor can anything 
which is not be present to anything else ; for what is not has 
no parts. 

Therefore we must deny of them not only the reality, but 
even the appearances of all the predicates which were formerly 
applied to them really or apparently, likeness and unlikeness, 
sameness and otherness, contact and separation, etc. 

The conclusion of the whole matter is, then, that, 
whether we assume that One is or that One is not, it itself 
and what is other than it, regarded both in themselves and 
in relation to one another, all are and are not, all appear 
and do not appear. 


202. And so it ends. No one has a word to say 
about this portentous result. If, however, we attend to 
the hints given in the course of the dialogue itself, we 
shall hardly be far wrong in drawing the following con- 
clusions from it. In the first place, the Megaric doctrine 
is refuted. If we postulate a One which is only one (as 
the Megarics did), we can say nothing whatever about it. 
Or if (as the Megarics also did) we identify One with 
Being, we shall have to predicate of it all sorts of incom- 
patible predicates. "Two statements'* (Sia-crol \oyoi) can 
be made about the One as well as everything else. 

On the other hand, the Sokratic theory has also been 
refuted in the early part of the dialogue, and that by argu- 
ments taken from the Megarics. It was based on the 
view that, though sensible things may partake in opposite 
forms, these forms themselves exclude one another. As 
that is untenable, we must try to find some other way in 
which things participate (a'XAo Set fyrelv 3> /xeraAa/x/Sai/ei}. 

The second part of the dialogue has shown once for all 
the impossibility of maintaining the isolation of the forms 
from one another. "The others" are just as hard to 
grasp as "the One." If we regard them abstractly, we 
can say nothing whatever about them ; while, if we regard 
them as being, we are compelled to ascribe contradictory 
predicates to them. In fact, the intelligible and incorporeal 
forms vanish under our hands just as the things of sense 
had done. It is clearly shown that we must now endeavour 
to understand in what sense the forms can participate in 
one another ; for all the difficulties of the Parmenides arise 
from the assumption that they cannot. 


The Sophist 

203. The Sophist is linked externally to the Theaetetus^ 
which is all the more remarkable that the evidence of style 
shows there was a distinct interval of time between the 
Sophist on the one hand and the Theaetetus and Parmenides 
on the other. The influence of Isokrates is strongly 
marked for the first time, especially in the avoidance of 
hiatus. In view of this interval of time, we shall be 
justified in looking for some real connexion between the 
dialogue and that of which it professes to be the sequel. 

Sokrates, Theodoros, and Theaitetos, with the younger 
Sokrates, his friend and later a member of the Academy, 
are supposed to meet again on the following day to con- 
tinue the discussion reported in the Theaetetus, but the 
fiction of the dialogue being read aloud at Megara is 
quietly dropped. The very title of the work is evidence 
of the growing coolness between Plato and the Megarics. 
Isokrates had already given the title of " Sophists" to the 
Sokratics generally, but more particularly to the "eristics," 
by whom he means mainly the Megarics, Plato adopts 
this way of speaking from Isokrates, and he also draws a 
hard-and-fast line between the Philosopher and the Sophist. 
That is made clear at the outset. A stranger from Elea is 
introduced, who is represented as a personal disciple of 
Parmenides and Zeno, and Sokrates at once professes 
alarm that he may prove to have a superhuman gift for 


274 LOGIC 

cross-examination. Theodoros reassures him, and says he 
is far too good a man for an eristic ; he is, indeed, a 
philosopher. Sokrates answers that it is hard to tell Philo- 
sophers from Sophists and Statesmen, and asks whether 
the Eleatics distinguished them. The Stranger replies that 
they did. 

Now Plato seems to speak to us more directly than ever 
before by the mouth of this Stranger, who, for that very 
reason, is anonymous ; and it seems, too, as if we were 
meant to understand once more that he claims to be the 
true successor of Parmenides, even though he is obliged 
to dissent from his central doctrine that " not being is 
not." What is this " not-being " which nevertheless is ? 
We shall find that it is identified with " the Other," and 
one of the few facts we know about the Megarics is that 
they said " What is is One and the Other is not"' x The 
name of Sophist is thus by implication applied to the 
Megarics, and it stuck to them. In fact, it more often 
means Megaric than not in the fourth century. We have 
heard of the "Sophist" Bryson and the "Sophist" Poly- 
xenos already ( 1 92). In Aristotle it is just the arguments 
of the Megarics that are technically called "sophisms," 
and it is with these he mainly deals in his course on 
fallacies. 2 If this is correct, I do not think it fanciful to 
suggest further that the reluctance of the Stranger to differ 
from his master Parmenides with regard to his central 
doctrine (241 d) is a hint of Plato's own attitude towards 
Sokrates at this time. 

Like several other dialogues, the Sophist appears to be 
made up of two wholly disparate sections bound together 
in an accidental way. It consists, as has been said, of a 
kernel and a shell. The shell is the attempt to find a 
definition of the Sophist by the method of division ; the 
kernel is a criticism of categories, especially that of " not 
being " (TO /4 &>). The ostensible link between the two 

1 Aristokles (af. Eus. P.E. xiv. 17,1; R.P. 289). 

2 The Tlepl cro^tcrrtKcoj/ IAe 


discussions is that the definition of the Sophist is found to 
imply the existence of " not being/' but that is by no 
means all. We find also that the reason why those who 
insist on the mere abstract unity of " what is " (TO oV) 
cannot advance beyond contradictory argument (avnXoyla) 
like that of the Parmenides^ is just that by so doing they 
have put it out of their power to divide any subject under 
discussion "according to its forms" or " kinds " (Kara yevtj, 
253 c-d). That is what the method of division aims at 
doing ; but it requires to be justified against those who 
deny that forms are a many, and that defence can only take 
the shape of a proof that " not being " (TO M Sv) is. Here, 
as in other cases, the real unity of the dialogue is left for 
us to discover if we can. 

204. It would be tedious to examine in detail the 
divisions by which the successive definitions of the Sophist 
are reached. They are not, of course, to be taken too 
seriously ; but neither, on the other hand, are they wholly 
without purpose. They are marked, in fact, by a certain 
not ill-humoured satire, the objects of which it will not be 
hard to guess after what has just been said. The Angler 
is first selected for definition, merely as an illustration of 
the method to be followed. That seems innocent enough; 
but it soon appears that the Sophist too is a fisher, a fisher 
of men, and this leads up to the definition of him as "a 
paid huntsman of rich and distinguished youths." That 
suggests another definition from the point of view of the 
art of exchange. He now appears as "a wholesale exporter 
of spiritual goods manufactured by others," though it is 
slyly added that he does sometimes dispose of his goods 
in the home market, and occasionally even manufactures 
them himself. Again, he may be looked on as a fighting 
man, whose weapons are short questions and answers ; or, 
again, he may fall under the art of sifting and purging. 
He purges the soul from beliefs that are a hindrance to 
knowledge, and especially from the ignorance which con- 
sists in thinking one knows what one does not know. 
Perhaps, however, we are doing the Sophist too high an 

276 LOGIC 

honour here, and this is a higher art than his. We may 
have been deceived by a resemblance. 

Obviously these last definitions do not apply to the 
great Sophists of the fifth century. Protagoras and Gorgias 
are always represented as averse to discussion by short 
questions and answers, and it is Sokrates who forces this 
method upon them. Again, the purging of the ignorance 
that consists in thinking one knows what one does not 
know is in the highest degree Sokratic. We are forced, 
then, to conclude that the persons aimed at are Sokratics, 
and the doubt expressed at the end of the discussion 
is an insinuation that they practised an imitation of 
the Sokratic method, though not always in the true 
Sokratic spirit. Once more it can hardly be doubtful who 
these are. 

205. The next section brings us to the real problem 
of the dialogue. We shall find that the Sophist's art is 
one that produces deceptive images and so gives rise to 
false judgements. On the other hand, the distinction of 
an image from the object imitated, and also the opposition 
of false judgement to true, imply that " what is not" in 
some sense is, and this Parmenides forbade us to assume. 
The argument proceeds as follows : 

We have given several accounts of the Sophist, but that 
shows there is something wrong with our method. His art is 
called by a single name, and there must, therefore, be some 
element which all these accounts of it have in common, and 
to which they all lead up. Now the account which seemed 
to point most clearly to this is the description of it as the art 
of Contradiction (avTt\oytKri). The Sophist professes to dis- 
pute on all things visible and invisible, in heaven and on earth, 
but it is impossible for one man really to understand all these 
things. Therefore the Sophist is a master of the Art of 
Appearance. He is like the painter who produces the appear- 
ance of solidity by lines and colours on a flat surface, and we 
may therefore call his art the Art of Imagery (eaScoAoTrouAoy). 
That art may be divided into two, that which produces an 
exact counterpart (eiKaa-rtia?) and that which produces an 
apparent likeness by deliberately altering the real proportions 
(<j>avravTiKri}. The Stranger is about to assign the Sophist's 


art to the latter when a pressing question of great difficulty 
emerges (232 a 236 d). 

How, indeed, can there be a deceptive image at all ? And 
further, how is it possible to say or think what is false, 
without which there can be no deceit ? In both cases we are 
forced to postulate that "what is not" is (vTroQe&Qai TO jmrj ov 
elvai), and that is just what Parmenides would not allow. If 
we say " is not," we must apply (7rpo<T<f>epiv) the words as a 
predicate to something. We cannot apply them to what is, 
and, if not, we cannot apply them to anything. But, if we are 
not speaking of anything, we are speaking of nothing, and are 
not in fact speaking at all. Nor can anything be applied 
(-TrpoorytyvGcrOai) as a predicate to " what is not." We cannot 
even say that it is one or many ; for number is, and we cannot 
predicate what is of what is not. But if "is not" can neither 
be subject or predicate, it is unutterable and unthinkable. 
Nay, we have no right to say that it is unutterable or unthink- 
able or even to call it " it " (239 a). 

Applying this to the Sophist, we find (i) that we cannot 
without contradiction speak of him as producing an image ; 
for, though an image is really an image, to be really an image 
is to be really unreal or really what is not (oimo? OVK oy). 
Nor (2) can we speak of his producing an unreal appearance 
(<l>dvTacrju.a) without contradiction; for that implies a judgement 
either that "what is" is not or that "what is not" is, and we 
.mave seen that such judgements are impossible. There is 
nothing for it, then, but to reconsider the dictum of Parmeni- 
des and to inquire whether we should not say that, in a certain 
sense, a what is not " is, and " what is " is not (241 d). 

A modern reader approaching this discussion for the 
first time is apt to think either that Plato is about to pro- 
pound a wanton paradox or that his mind is obsessed by 
the spectre of some fantastic <c metaphysical " conception 
of Non-being. That is, firstly, because he is using 
the language of his time, a language which he did not 
invent and for which he is not responsible. If he had 
been writing for us, he would no doubt have formulated 
the problem in another way. As it was, the Megarics had 
inherited from Parmenides the doctrine that "what is 
not" is not (a doctrine which, in the mouth of its author, 
had a purely cosmological significance), and they had 

278 LOGIC 

imported it into Dialectic, with the result that they were 
led to deny the possibility of significant negation. In the 
second place, the extreme simplicity with which the problem 
is stated is disconcerting to the modern mind. That is 
characteristic of Greek philosophy as a whole, and is one 
of the things that makes it worthy of study. There is 
nothing like stating difficulties in their baldest form to 
ensure that they will not be evaded. The modern reader 
would feel no difficulty if Plato had announced a discus- 
sion of the possibility of significant negative judgements, 
and that, as a matter of fact, is the subject of this dialogue. 1 
It is a good thing, however, to study it in its simplest form 
and stripped of conventional terminology. 

206. In reality, the Stranger proceeds, the reason why 
we find such difficulties in " not being " is just that we do 
not know what is meant by "being." Earlier philosophers 
have not taken the pains to think out clearly the import 
of certain elementary terms, the meaning of which appears 
to be obvious, but is really very far from being so. That 
is why they have only been able to tell fairy tales. Some 
say the things that are (ret 6Wa) are two or three or some 
other number. Others maintain that what is is one ; 
others, again, seek to combine these views. But no one has 
asked what we mean by saying of anything that it is. 
This is shown by a criticism of the Pythagoreans, who 
said things were two, and of the Eleatics, who said they 
were one. 

If all things are two (e.g. hot and cold), how is the "being" 
which this implies related to the two ? Either it must be a 
third thing besides them, or it must be identified with one of 
them, in which case the other would not be, Or, if we say 
that "being" is true of both in the same way, they will be 
one and not two (243 d 244 a). 

If all things are one, then "being" and "one" are the 
same, and only two names for the same thing. But, apart 
from the absurdity of having two names for the same thing, 

1 It is precisely the problem discussed in Bosanquet's Logic, Bk. L 
chap, vii., which will be found to throw light on the Sophist* 


how can there be a name at all ? If the name is other than 
the thing, they are two and not one, so that, if all things are 
one, there can only be a name which is a name of nothing, 
or the thing itself will be a name, and its name the name of a 
name (244 b-d). 

But they also say that the one which is (TO ov <lv) is a 
whole. But a whole has parts and is therefore other than 
one, which as such is indivisible. If, then, "what is" is a 
whole, it is a many. On the other hand, if it is not a whole, 
it is not the whole of what is, and it can neither come into 
being nor be ; for what comes into being or is comes into 
being or is as a whole (244 d 245 d). 

This is, of course, a summary of certain arguments in 
the ParmenideSy and has a similar purpose. It is as hard to 
grasp the meaning of is as it is to grasp the meaning of is 
not. The difficulty is even greater when we turn from the 
number of what is to its nature. 

207. With regard to this there is a regular battle of 
the gods and giants between philosophers. Some identify 
reality or being (ova-la) with body, that which admits of 
impact and contact (o Trapeze* irpoo-ftoXrjv mi eiraffiv Tiva), 
while others say that true being consists of certain 
intelligible and incorporeal forms or figures (I/OJ/TCI arra 
tcai ao-c^uara ?&;), while everything corporeal is only a 
stream of becoming (jpepoimevq yevea-ti). 

We must pause here and ask to whom the Stranger is 
referring ; for this is one of the most pressing questions 
in the history of Greek philosophy. In the first place, it 
must be observed that the philosophers now under dis- 
cussion are spoken of as if they belonged to a past genera- 
tion. It can hardly be correct to suppose that the school 
of Demokritos are intended by the " earth-born " (yyyeveii). 
Demokritos, who asserted the reality of the void, could 
not be spoken of as making impact and contact the test 
of being. We have seen, however, that the doctrine of 
Parmenides paved the way for materialism, and that 
Melissos, who was a very important figure in the latter 
part of the fifth century, definitely taught a materialistic 
monism (68). As to the "friends of the forms" 

280 LOGIC 

(elSwv <pi\oi), of whom Plato speaks with such aloofness by 
the mouth of the Stranger, if our general view of the 
doctrine of forms is correct, we have seen that there is no 
difficulty in identifying them with the later Pythagoreans. 1 
At any rate, they can hardly be the Megarics, as is often 
supposed ; for they rejected the plurality of forms alto- 
gether, and identified the One and the Good ( 129). 
It is worthy of note that the Stranger speaks of them as 
persons whom he understands, " thanks to his intimacy 
with them " (Sta avvqdeiav), and that suggests they were to 
be found in Italy. The language in which their doctrine 
is described is just that of the first part of the Phaedo^ and 
they may therefore be identified with the "we" of that 

208. The corporealists are hard to deal with ; but, if 
we imagine them for the moment to be more reasonable 
than they are, we may get them to admit that by reality 
or being (ova-la) they in fact mean force 

They must admit that there is such a thing as a mortal 
animal, and therefore as an animate body, and therefore as a 
soul. They must further admit that a soul may be good or 
bad, wise or foolish, and therefore that goodness and wisdom, 
the presence or absence of which make it one or the other, 
are. Very likely they may say that the soul is body, but they 
will hardly say that goodness or wisdom are bodies (though 
it is to be feared the real earthborn would). But, once they 
admit that a single incorporeal thing is, they must accept a 
definition of being which will apply equally to it Perhaps 
they may accept as a definition of what is that it is anything 
that has the least power of acting and being acted upon, that, 
in fact, being is force (246 e 247 e). 

It is to be observed that the Stranger does not put this 
definition forward as one satisfactory to himself. Indeed, 
he says expressly that we shall very likely take a different 
view later. 

If we turn now to those superior persons, the " friends 

*As we have seen (p. 91, n, i) this identification is made without 
hesitation by Proclus, and is presumably the Academic tradition. 


of the forms/' we may expect them to be more tractable, 
and more ready to admit that what is is what can act and 
be acted upon. As a matter of fact, however, we shall 
find them even less amenable to argument than our 
reformed corporealists. They remain in the sky and do 
not answer us at all, though the Stranger knows from his 
intimacy with them that they regard us with contempt. 
They will not ascribe any kind of motion at all to reality 
or being (own'a), and therefore they will not speak of 
acting or being acted upon in connexion with it. 

The "friends of the forms" distinguish being (ova-la) from 
becoming (ytWori?) and say that our souls participate in 
constant being by means of thought, and our bodies in variable 
becoming by means of sense. But this participation surely 
implies that being has a power of acting and being acted 
upon ; for the thought that knows being must, in so doing, 
either act or be acted upon or both, and the being that 
thought knows must accordingly either act or be acted upon 
or both. 

To this we may suppose them to reply that being is 
constant and immovable, and cannot therefore either act or be 
acted upon. But they must admit that we know being, and 
knowledge implies soul, and soul implies life and motion. If 
these are excluded from being and referred to becoming, there 
can be no knowledge at all. It is equally true, however, that 
being would be unknowable if it were only variable and in 
motion ; for knowledge implies constancy, and that implies 
rest (248 a 249 d). 

We have not been able to get any answer out of the 
" friends of the forms " ; but our discussion with them has 
suggested that knowledge is impossible unless being is 
both in motion and at rest. But, as motion and rest are 
opposites, they cannot be united. On the other hand, 
they both are, and therefore being must be a third thing 
over and above them. From this it follows that being 
per se is neither at rest nor in motion. What are we to 
make of this ? We see, at any rate, that it is just as hard 
to say what is meant by is as to say what is meant by is not, 
and this gives us a ray of hope. If we can only discover 

282 LOGIC 

what is means, the other difficulty may be got rid of at the 
same time. 

209. We must start from the fact that, when we speak 
about a thing, we not only name it, but apply many other 
names to it. When we speak about a man, for instance, 
we apply to him the names of colours, forms, sizes, virtues 
and so forth. Of course there are youthful logic-choppers 
and elderly amateurs (Antisthenes ?) who say we have no 
right to do this. Man is man, and good is good ; but, if 
we say " the man is good," we are confusing the One and 
the Many. Such theories are sufficiently refuted by the 
fact that they cannot be stated without contradiction. 
Those who forbid us to say that A is B in virtue of A's 
" participation in being affected by 1 " B (252 b) have them- 
selves to use such terms as "is," "apart from," "from 
others," " by itself," and thus carry about with them an 
inner voice that refutes their theory. 

We must say (i) that all things are incapable of participating 
in one another, or (2) that all things are capable of participating 
in one another, or (3) that some things are capable of partici- 
pating in one another and others are not. In the first case, rest 
and motion cannot partici pate in being, and so cannot be. That 
makes havoc of all the theories we have considered hitherto. 
In the second case, it will be possible for motion to rest and 
for rest to move. Only the third case is left, namely, that 
some things can participate in one another and others cannot 
(252 e). 

We shall find that these simple considerations suggest 
the solution of the difficulty we have been dealing with. 

This solution is briefly that is and is not have no meaning 
except in judgements or predications (\6yoi). In one 
sense, this doctrine is not new. In the Phaedo Plato 
made Sokrates formulate the method of seeking for truth 
in judgements (ev rolg Xo-yot?), and there too we have the 
terminology which represents the subject as " partaking " 

1 The phrase KQWWVIV. TraOrjf^aros lre/oou is derived from the use of 
TrtTTovOevat, to express the relation of a subject to a predicate. Cf. Farm. 


in the predicate, and also the way of speaking according 
to which the subject " is affected by" (-TreVoj/flei/) the 
predicate. 1 What is new here is that, whereas in the 
Phaedo it is the particular things of sense that " partake 
in " the forms, we are now discussing the participa- 
tion of the forms or "kinds" (7^) with one another. 
The need for such discussion has been shown in the 
Parmenides ( 194, 199). It is to be observed further 
that these forms or "kinds" of which we are now 
speaking are just the common predicates (KOIVO) of the 
Theaetetus ( I 86). We may say, if we like, that these are 
the Platonic forms as distinct from the Pythagorean or 
the Sokratic. 

210. We have found that some forms or kinds will 
participate in one another and others will not, just as some 
letters will go with one another and others will not. The 
vowels, in particular, pervade all combinations of letters, 
so that without a vowel there cannot be any combination 
at all. In the same way, some notes in the octave are 
concordant and others are not. In these two cases we 
have the arts of Grammar and Music to direct us, and so 
we require an art which will show us what forms will 
harmonise with one another and what forms will not, and 
especially whether there are any kinds which (like the 
vowels) pervade all combinations and disjunctions (e.g. is 
and is not}. That is just the art of Dialectic, and the man 
who possesses that will be able to distinguish what forms 
can enter into combination and what will not. 

In particular, he will be able to distinguish (i) a single form 
pervading many single and separate things, (2) many forms 
distinct from one another but comprehended from without by 
one, (3) a single form pervading in turn many such wholes 
and binding them together in one, while many other forms 
are quite separate and apart from it (253 d). 

This passage gives us the foundation of Plato's Logic. 
The following points in it should be noted : 

(a) He distinguishes clearly between (i) genus and (2) 


284 LOGIC 

species, though he uses the terms form and kind (eTSof, 
iSea, <yeVo9) indifferently of both. 

() The single forms described under (3) are the " highest 
kinds " (/meyio-ra yeV;), such as Being, Rest, and Motion. 
These are all of them " manners of participation," or, as 
Aristotle called them, " forms of predication" (cr^/xara 
rfjs Kartiyoplai). They have no meaning except in a 

(c) In the Phaedo the question was what particular things 
admit a given form as their predicate ; here the question 
is one of the compatibility or incompatibility of the 
"highest kinds" or forms with one another. Is it possible 
for any of these to be predicated of one another ; and, if 
so, which can be so predicated and which can not ? 

(*/) As Being is only one of the categories, though the 
most pervasive of all, it has no meaning except as entering 
into a judgement. By itself the word "is" means 
nothing ; it is only the bond that unites a subject to a 
predicate. We may put this by saying that Plato for the 
first time discovered " the ambiguity of the copula," 
though, for reasons which will appear, he would certainly 
not have put the thing in that way. 

211. To avoid confusion, let us select only a few of 
the " highest kinds " (/xeyja-ra yevrj} and consider (i) their 
nature, and (2) which combine with which and to what 
extent. In this way we may be able to discover some 
sense in which we may safely say that there really is such 
a thing as <c not being." To begin with, Rest and 
Motion exclude one another, but both of them are, and 
therefore combine with Being. That gives us three kinds, 
but each of the three is other than the other two and the 
same as itself. That gives us a fourth and a fifth kind, 
Same and Other; for we cannot identify these with any of 
the first three. 1 

For (i) if we identify either Rest or Motion with any 
common predicate of both, then it will be predicable of the 

J Cf. Theaet. 185 a sq. (above, p. 247), 


other, so that Motion will rest or Rest will move. But Same 
and Other are common predicates of Rest and Motion, 
therefore neither Rest nor Motion can be identified with 
Same or Other. Again, (2) if we identify Being and Same, 
then, as Rest and Motion both are^ they will be the same. 
Lastly, (3) we cannot identify Being and Other ; for Other 
is essentially (rovro 6Vep ecrrlv) relative (Trpo? erepov) and 
Being is absolute (/ca(9' cwro). Therefore Other is a fifth 
kind (255 a-d). 

Now Other pervades all the rest, just like Same and 
Being ; for each of them is the same as itself and other 
than the rest, and this amounts to saying that each of them 
is itself and is not any of the others. 

Thus Motion, being other than Rest, is not Rest, but it is 
Motion. Motion, being other than Same, is not Same, but it 
is the same as itself. (We must not mind the apparent 
contradiction. If we had not shown that Motion and Rest 
exclude one another, we might even have to say that Motion 
was at rest.) Again, Motion, being other than Other, is 
Other in a sense and is not Other in a sense. Lastly, Motion, 
being other than Being, is not Being, but it is Being because 
they all partake in Being. Motion, then, is really both 
Not being and Being, and the same thing will apply to all the 
other kinds, since each of them is other than Being and each 
of them is (255 e 256 e). 

We may say, then, that each of the kinds, in virtue of its 
otherness, has much Being and infinite Not being. And, 
as Being itself is other than all the rest, we must say that 
Being is not just as many times as there are other things, 
and they are innumerable. Not being these, it is just 
itself, but it is not the rest innumerable times. 

212. But this Not being which we have discovered is 
not the opposite of Being (like the Not being Parmenides 
spoke of). The negative term (airotfHxris) produced by 
prefixing " not " to a word only signifies something other 
than the word which follows the negative, or rather than 
the thing that word denotes. Now otherness is subdivided 
into as many parts as knowledge, so, just as there are 
many sciences and arts with names of their own, the parts 

286 LOGIC 

of otherness will have names of their own. The part of 
otherness opposed (avTiTi6ejut.evov) to the beautiful is the 
not-beautiful, which is not other than anything else but 
beauty, and the not-beautiful is just as much as beauty, 
and so of the not-great, the not-just, and so forth. It is 
in this combination with a particular part of Being that 
Not being really is ; it is " not being so-and-so," and it is 
just as much as what it is not. We need not trouble 
ourselves further, then, about the question whethe^ Not 
being as the opposite of Being can be thought or spoken 
of or not. In the sense we have now given it, it certainly 
is and is all-pervasive. It is merely childish to separate 
Being from Not being, and to argue that a thing must 
either be or not be. The two forms are inseparably bound 
up with one another, and this is what makes rational 
Speech possible (Sta yap TTJV aXX?JXcov T&V eiSwv KOiv&vlav 6 
Aoyo? yeyovev JJJMV 259 e). 

What has been proved so far is (i) that everything that 
is positively determined is also negatively determined, and 
(2) that negative terms are an expression of reality 
(SrjXdo/mara -nf? ovcr/a?). It has been shown further, (3) that 
the reality expressed by a negative term is not the contrary 
of the corresponding positive term, but its contradictory. 
On the other hand, it has been shown (4) that, as the 
negative term must always be understood in relation to 
the corresponding positive, the reality it expresses is always 
a particular part of reality, so that "not-great, 77 for 
instance, does not include "beautiful" or "just," but only 

213. In the course of the foregoing discussion the 
remark was thrown out that we have found the Not being 
which was necessary to justify our account of the Sophist. 
This is not explained further, but the point is quite simple. 
We called him an image-maker, and he replied that there 
was no such thing as an image, since an image is really not 
real. We now see that there is nothing in this objection ; 
for the art of image-making, like all other arts, includes a 
part of Being and a part of Not being. The image is not 


the reality, indeed, and the reality is not the image, but 
that involves no difficulty. We are dealing with a par- 
ticular art, that of Image-making, and in it "not real" has 
a perfectly definite and positive signification. The " not 
real" is not the unreal, but just the image, which is quite 
as much as that of which it is the image. 

Even admitting this, however, the Sophist may still say 
that it is impossible to say or think what is false. Though 
we have shown that Not being z'j, or in other words that it 
combines with Being, we have not shown that it combines 
with speech. But, unless it does so, falsehood is impossible, 
and so therefore is deceit. We must, therefore, scrutinise 
carefully (i) speech (Aoyo?), (2) judgement (5o'a), and 
(3) appearance (jcpavracrla), with the view of seeing whether 
Not being and consequently falsehood can enter into them 
or not. 

We must begin, as we did in the case of letters, by con- 
sidering whether all words combine with one another, or 
whether some will and some will not. There are two kinds 
of words that are expressions of reality (cfyAco/xara ry? oucrw), 
nouns (oWyuara) and verbs (prjjmaTa)- The latter express 
action or inaction or the reality of being or not being (i.e. the 
reality expressed by a positive or negative term) ; the former 
express the agent^ or what is or is not so-and-so. A statement 
(Ao'yo?) cannot consist of nouns alone or of verbs alone ; the 
very simplest must have one of each, e.g. "man learns." 
Further, every statement must be "of some one or something" 
(rtvo$ e?j/cu)> an( i it must have a certain quality (TTOIOV TWO. 
ewx*) 5 i.e. it must express something which is or becomes 
in the present, past, or future (rcoy OVTOOV rj yiyvoju-evcov rj 
yeyovorcov >? /xeAXoVrcov). 1 Now let us make a simple experi- 
ment. If I say "Theaitetos is sitting," that is a statement 
which is " of Theaitetos," and it has the quality of expressing 
something which really is at the present moment. But, if I 
say " Theaitctos, to whom I am talking at the present moment 
(vvv)> is flying," that is also a statement which is " of Theai- 
tetos," but it has the quality of saying something of him which, 

1 That "quality" really means tense seems to follow from the context, 
and especially from the emphasis on "to whom I am talking at the present 
moment" in the illustration which follows. 

288 LOGIC 

though expressing a real action, is something other than what 
is real with regard to Theaitetos at the present moment. It is, 
therefore, possible to speak of what is not as being, and that 
is what we mean by falsehood (261 d 263 d). 

In fact, what we call truth and falsehood are not to be 
found in terms, whether positive or negative, but only 
in the proposition, which is a copulation (o-vjmTrXoK^) of 

214. It will be observed that significant negative judge- 
ment is explained as the affirmation of a negative predicate 
(a7ro'<jt>a<w), but it would be altogether wrong to identify 
this with what Aristotle calls an "indefinite" predicate 
(aopta-Tov p/m), that is, a predicate which may be truly 
predicated of everything alike, whether existent or non- 
existent. In the present case, for instance, " is sitting " 
excludes every other form of Rest, and therefore "is 
sitting" implies the negative judgements "is not lying/' 
"is not standing," and whatever other forms of Rest 
there may be. In the second place, " is sitting " excludes 
all the forms of Motion, which cannot have any com- 
munion with Rest, and therefore implies the negative 
judgements "is not walking," "is not running," "is not 
flying." The significance of the negative judgement 
depends, in fact, on the system of kinds and forms to 
which it refers, what we should call a "universe of dis- 
course." Plato held that there was a perfectly definite 
number of such forms in each kind, which it is the 
business of the dialectician to discover. That is why he 
insists .that "not being" is subdivided into as many sub- 
divisions as the arts, and that each "part" of "not being" 
can be understood only in relation to the corresponding 
"part" of "being." The negative predicate "is not 
flying" does not include "is beautiful" or "is just." 

In the present case, the predicate " is flying " expresses 
a real form of action, a real form of the kind Motion, and 
it is "of Theaitetos," who is a real agent. The reason why 
the statement "Theaitetos is flying" is not true is just 
that, at the -present moment (wv) 9 Theaitetos " is sitting," 


and that predicate excludes " is flying." It does not exclude 
" was flying " or " will be flying/ 1 and that is why we must 
attend to the "quality" of the statement. 1 

2 1 5. But, if it is possible to say what is false, it is also 
possible to think what is false ; for thought only differs from 
speech in this respect, that it is cc the conversation of the 
soul with itself taking place without voice," while speech 
is " the vocal stream issuing from the soul through the 
lips." Now we know that positive and negative predica- 
tion ((pdcris and aTro'^acn?) are found in speech, and, when 
the same things occur silently in the soul, we call them 
judgement (<W|a). Again, when affirmation and negation 
take place in the soul, not in virtue of its own activity, but 
through the agency of sensation, we call that appearance 
((pavrao-ia). It follows that, as thought (Sidvoia) is mental 
speech, and judgement (<Joa) is "the completion of 
thought," and appearance (jpavrao-la) is a mixture of sensa- 
tion and judgement, the truth and falsehood which are 
possible in speech will also be possible in judgement and 
in appearance. 

Now that he has shown the possibility of false judgement 
and false appearance, the Stranger goes on to give his final 
definition of the Sophist. That is of no particular import- 
ance for us here, though we may note some interesting 
points. Of these the most significant is the way in which 
advantage is taken of the division of productive art into 
divine and human to assert in impressive language the 
doctrine that what we call natural objects are the work of 
God and not of Nature or of Chance. We shall see 
presently that this thought was occupying Plato's mind at 
the time, and that he was already trying to work out a 
rational justification of theism. 

1 Most commentators understand by " quality " the truth or falsehood 
of the statement, but that would make the argument puerile. There is 
no point in asking how we know that Theaitetos " is sitting " now. We 
see him, of course. 



The Statesman 

2 1 6. The dialogue entitled the Statesman (IIoAm/co?) 
is in form a sequel to the Sophist. The characters are the 
same and the leading part is still taken by the Eleatic 
Stranger. There is no reason to suppose that the two 
dialogues are separated by any considerable interval of 

The discussion begins by an attempt to find the defini- 
tion of the Statesman by the method of division, and it is 
easier to trace the connexion of this with the principal 
theme of the dialogue than it was in the case of the Sophist. 
The first definition we reach represents the King as the 
Shepherd of Men, as he is already called in Homer. 
There is good reason for believing that this was the 
Pythagorean view. The King to them was an <c image " 
of God upon earth ; for God was the shepherd of the 
world. 1 This is, in fact, the theocratic ideal of kingship. 
The Eleatic Stranger points out, however, that it rests on 
a confusion between God and man, and could only be 
realised if God were in person our ruler. That is the 
point of the myth related by the Stranger. The course 
of the world was once directed by God himself, but we are 
not living in that age. There are seasons when the captain 
of the world-ship (a Pythagorean conception 2 ) retires to 

1 See Campbell's Introduction to the Statesman, p. xxv sq. 
*E. Gr. ?/5. 2 p. 342. 


his conning-tower and leaves the ship to itself. At those 
times the world goes round in the opposite direction to 
that which God had given it, and all natural processes 
are reversed (an idea which may have been suggested by 
Empedokles). We are living in one of these periods, and 
there can be no question for us of a divine ruler. There 
is a curious hint that, after all, the ideal of mankind as a 
flock or a herd fed by the hand of God may not be the 
highest. If the men of those days, who had no need to 
take thought for the morrow, and who found everything 
bountifully provided for them without any labour on their 
part, spent their time in gathering wisdom, and made use 
of their power to communicate with the beasts in the 
interests of philosophy, then indeed they were happier 
than we are. But if they and the beasts spent their time 
in telling fables to each other such as have been handed 
down by tradition to our own days, it is not hard to form 
a judgement as to that either (272 c). This passage is 
very important. It is plain that the theocratic ideal of 
the Pythagoreans had little attraction for Plato. He did 
not think we could get rid of problems by simplifying them 
out of existence. 

217. Let us turn, then, from the divine ruler to the 
human. He will not be the feeder of his flock, but only 
its tender (275 e). He will have complete knowledge 
of what is good for his subjects, and he will secure it for 
them with or without their consent, just as the doctor who 
knows what is good for the body will cure his patients 
whether they like it or not. He will have no need of 
laws. No law can take account of the infinite variety of 
particular cases ; it can only lay down certain principles in 
a rough and ready way. If the ruler were able to attend 
to every case in person, and if he could always be present, 
it would be absurd for him to trammel himself with laws. 
If he had to go away for a time, he would no doubt make 
laws to guide his subjects in his absence, just as a doctor 
might leave behind him written instructions for his patient. 
But, when the doctor came back, it would be ridiculous for 


him to insist on keeping to these instructions. He would 
feel quite free to alter the treatment if he saw fit. In the 
same way, if the philosopher king were ever to appear on 
earth (as he may have done in the past), there would be no 
need of laws. At present there is no appearance of his 
return, so we must do as well as we can without him. We 
must try to frame laws as nearly as possible in accordance 
with what he would approve, and we must insist upon their 
being scrupulously observed. If men found they were 
being badly treated by the practitioners of the arts of 
medicine and navigation, they would insist upon a code of 
rules for these arts being drawn up, and upon all trans- 
gressions of these being punished, and that is the true 
place of law in the state. It is only a makeshift (Seurepos 
7rA<w) ; but, as things are, it is indispensable. It is in 
this way that Plato deals with the philosopher king of the 
Republic. His rule is still the ideal, but there is no 
immediate prospect of it being realised. The use of such 
an ideal is nevertheless very great. In the first place, it 
gives us a standard by which we can judge existing or 
possible institutions, and in the second place, it will save 
us from the mistake of attaching too high a value to these, 
and refusing in consequence to contemplate any alteration 
of them. The true point of view from which to regard 
existing laws and institutions is to look on them as more or 
less tolerable expedients. They are all alike open to criti- 
cism when compared with something higher, and ultimately 
with the rule of the philosopher king. We may say, then, 
if we please, that the purpose of the Statesman is to deter- 
mine the provinces of realism and idealism in politics. 
We must not put the ideal too high, as the theocratic ideal 
did, but we may make it as high as we please, so long as 
we take account of human nature. The analogy of the 
beasts of the field is inapplicable to mankind. 

218. Plato goes on to give a classification of constitu- 
tions from this point of view, and, as might be expected, 
it is quite different from that of the Republic. There are 
six constitutions altogether, the rule of the philosopher 


king being excluded as hors concours. The basis of division 
is twofold. The rulers may be one, few, or many, and they 
may rule according to law or lawlessly. Of the legal con- 
stitutions, kingship comes first, aristocracy second, and 
democracy third ; for the possibility of political knowledge 
is inversely proportional to the number of rulers. But, 
when we come to the lawless constitutions, the order is 
reversed. There is only one name for a constitutional 
and a lawless democracy, but they are quite different in 
principle. Of all possible constitutions democracy can do 
the least good and the least harm, so that, while a consti- 
tutional democracy is inferior to aristocracy and still more 
to constitutional monarchy, even a lawless democracy is 
far superior to a lawless oligarchy, and still more to a 
lawless tyranny. Such is the view of Plato, but it would 
be very hard to imagine Sokrates accepting any such doc- 
trine. Even the Periklean democracy is not harshly 
treated. It is, of course, a lawless democracy, but it is not 
condemned so bitterly as it was in the Gorgias and the 
Republic. If it cannot do much good, it does relatively 
little mischief. The legal democracy is more or less the 
Athenian democracy of Plato's own time, and is placed 
just below true aristocracy. All this is quite in keeping 
with what we have learnt as to Plato's political upbringing 
and experience ( 158), and it agrees very well with what 
he says about his political attitude in Epistle vii. It was 
impossible to maintain the Sokratic condemnation of all 
democracy after the events which marked the end of the 
fifth century. 

But that is not all. Plato does not insist in a doc- 
trinaire fashion on any rigid classification of constitutions. 
One of the chief functions of the true ruler is just to unite 
the various elements in the state, as the weaver unites the 
warp and the woof of his web, and there is room for a 
number of mixed constitutions as well as for the six types 
already described. In the Laws Plato's final conclusion is 
that, as things are, and in the absence of the philosopher 
king, the best constitution will be a combination of legal 


kingship with legal democracy. 1 He is thus able to take 
an extremely practical view of political questions, and he is 
able to do so without abating one jot of his idealism. 
That is where he goes beyond Sokrates, whose political 
teaching had not, we have seen (145), been an unmixed 
blessing to his country. 

Plato and Dionysios. 

219. Plato's political teaching in the Academy had an 
enormous influence through his pupils ; for the foundations 
of Hellenistic civilisation were mainly laid by them. His 
personal intervention in the politics of the Hellenic nation, 
which was already coming into being, was in some ways a 
failure, as the world counts failure. He expected it to be 
so, and he entered upon it with great misgiving ; but it 
seemed worth trying, nevertheless. It was just possible 
that he should succeed, and friends of his who were in a 
position to form a judgement were confident that he would, 
so he felt unable to shirk the task offered to him. To 
decline would have been treason to philosophy (Ep. vii. 
328 e). If he had succeeded, the course of European 
history would have been altered, and we shall see that his 
failure was due to causes beyond his control. 

In 367 B.C. Dionysios I. of Syracuse died at the age of 
sixty-three, after a reign of thirty-eight years. He was in 
many ways a great man, but he had failed in the main 
purpose of his life, which was to drive the Carthaginians 
from Sicily. He had been defeated by Hanno the year 
before his death, and a peace was now concluded on the 
basis of the status quo ante bellum. His successor, Diony- 
sios II., was nearly thirty years old, but he was quite unfit 

1 In the Laws the best constitution is a mean between Persian monarchy 
and Athenian democracy (756 e). Apparently Plato would have been an 
admirer of the British Constitution. It is also worthy of note that his 
ideal is not very unlike that of the speech of Perikles in Thucydides, and 
is just what might be expected of the stepson of Pyrilampes, That does 
not, of course, imply approval of Periklean democracy with Perikles left 
out. The illustration from the art of weaving is common to the Statesman 
and the Laws (734 e /#?.) 


to take up the reins of government His father had 
always been jealous of sharing his power with anyone, and 
had even sent his ablest minister, Philistos the historian, 
into exile at Adria, near the mouth of the Po. For the 
same reason he had purposely kept his son at a distance 
from all public affairs, and encouraged him to find amuse- 
ment in such pursuits as amateur carpentry and turning. 
The young man was not, we are told, without natural 
gifts, and it seemed to Dion, who was his father's brother- 
in-law and a devoted admirer of Plato, that something 
might still be made of him. It was too late to send him 
to the Academy at Athens, which by this time was the 
recognised institution for the training of rulers and princes, 
so Dion conceived the scheme of bringing Plato, now sixty 
years old, to Syracuse. There was nothing in the least 
chimerical in the project, and the problems Syracuse had to 
face made it essential that she should have an enlightened 
ruler. The great question of the day was once more how 
Hellenism could maintain itself against the pressure of 
Persia on the one side and Carthage on the other, and far- 
sighted statesmen saw clearly that the only hope lay in 
taking the offensive. We hear most, as is natural, of 
Persia. The conditions imposed by the King's Peace^of 
387 B.C., which left the Greek cities of Asia under Persian 
rule, were humiliating and intolerable. That side of the 
problem was successfully dealt with later by Alexander, 
and it was from the Academy that he derived his inspira- 
tion ; l but the situation in Sicily was quite as serious. 
The Carthaginian question was only another aspect of the 
Persian question, and it is at least an instructive tradition 
that represents the battles of Salamis and Himera as having 
been fought on the same day. 2 

^Plut. adv. Co!. Iiz6d. Delios of Ephesos, an associate (!TCU/>OS) of 
Plato, was sent to Alexander by the Hellenes who lived in Asia, and did 
most to enname him and stir him up to engage in war with the barbarians. 

2 It is interesting to note that the struggle between Hellenes and 
Semites had also been going on in Cyprus, the other great *' meeting- 
place of races." Isokrates played a similar part there to that which Plato 
played in Sicily, in his own way, of course. 


220. Plato refused, however, to let things be rushed. 
Dionysios had a great deal of ground to make up, and 
it was necessary for him to go through a serious course 
of higher study before he could be trusted to make 
even a beginning with schemes of reform and liberation. 1 
Unfortunately he was rather old for this. According to 
Plato's own principles, he ought to have begun these studies 
at the age of twenty, so it was natural enough that, after the 
first enthusiasm had passed, he should feel them irksome. 
That was the opportunity of the opposition who still clung 
to the principles of the elder Dionysios. Philistos (or, as 
Plato calls him, Philistides) had been recalled from exile, 
and he set himself at once to undermine the influence 
of Dion and Plato. The somewhat masterful and haughty 
temperament of Dion also played into his hands, and 
it was not hard to persuade Dionysios that his kins- 
man was taking too much upon himself. Only four 
months after Plato's arrival Dion was banished, and Plato 
saw it was all over with the project of reform. On the 
other hand, Dionysios had no idea of losing Plato, to 
whom he had become deeply attached. He had, in fact, 
been jealous of Dion's intimacy with him, and hoped to 
have him more to himself now Dion was out of the way. 
It was not to be expected that Plato would give up his 
friend, however, and he pressed his claim in season and 
out of season. A situation which threatened to become 
impossible was ended by the outbreak of war. Dionysios 

1 Grote thinks Plato was wrong here, but that seems very doubtful. 
If he was not to give Dionysios a regular training like that of the 
Academy, what was the use of his coming to Syracuse at all ? Possibly 
the men of those days believed too much in science, but their belief in it 
was perfectly sincere. Prof. Bury's view is even more remarkable. He 
thinks (vol. ii. p. 247) that Plato should have contented himself "with 
inculcating the general principles which he has expounded with such 
charm in the Republic" in which case "Dionysius would in all likelihood 
have attempted to create at Syracuse a dim adumbration of the ideal 
state " ! In that case, we may add, the Carthaginians would have 
annexed Syracuse. Plato was no Utopian dreamer, and the notion that 
he proposed to introduce the arrangements of the Republic at Syracuse 
(of all places) is quite unsupported by any sort of evidence. 


had to interrupt his studies, and Plato was free to return 
to Athens. The understanding was that at the conclusion 
of the war Dion should be restored to his old position, 
and that then Plato would return. On his way home he 
visited Archytas at Taras. 

221. It is not very likely that Dionysios was sincere 
in his promise to become reconciled to Dion, but he 
was determined to get Plato back at all costs. He tried 
to carry on his mathematical studies in his absence, and 
made the subject quite fashionable at court. At first 
Plato declined to return unless Dion was reinstated, but 
he was urgently entreated to do so by Dion himself and by 
Archytas, the most successful statesman of the day. He 
ought certainly to have been a good judge of the situation, 
and he assured Plato that Dionysios was really enthusiastic 
about philosophy, and that everything would now go 
smoothly. With great reluctance Plato accordingly made 
up his mind (361 B.C.) to "recross Charybdis" (Ep. vii. 
345 e) ; but he soon discovered that Dionysios had not 
the slightest intention of doing anything for Dion, and 
a breach became inevitable. Plato wished to go home, 
but Dionysios would not let him. No ship captain would 
venture to take him as a passenger in the circumstances, 
and he had to wait a whole year. At last a violent 
quarrel broke out on the occasion of a military revolt. 
Dionysios made Herakleides, one of his officers, respon- 
sible for it, and Plato with great difficulty got him off* 1 
Dionysios could not forgive the way in which he had been 
shamed into an act of clemency, and bitterly reproached 
Plato with having hindered him in the work of reform 
and the liberation of the Hellenic cities under Carthaginian 
rule. Instead of that he had made him learn geometry ! 
Plato was excluded from the court and practically kept a 
prisoner, until, on the intercession of Archytas, he was at 
last allowed to return to Athens (360 B.C.). Even then 

1 We gather from the Epistles that Plato was very unpopular with 
the mercenary troops. These wild Keltic warriors knew very well that 
if Plato had his way their day was over. 


there was no final breach. Dionysios kept writing to 
Athens for explanations of difficult points, and Plato 
answered him. He even wrote a book, much to Plato's 
annoyance, in which he professed to disclose the Platonic 
philosophy. It is clear that Archytas and Dion were not 
wrong in believing he had some natural gifts, but they 
had not been cultivated early enough. He was vain and 
petulant, no doubt, but his attachment to Plato was 
obviously sincere, and we cannot help feeling a little 
sorry for him, when we remember what he might have 
been if his father had given him a chance when he was 
young enough to profit by it. 1 

222. At this point Plato's personal responsibility for 
the affairs of Syracuse ceases, but Dion was still to be 
reckoned with. He was not the sort of man to wait for 
ever, and he began to collect adherents all over Hellas. 
He had determined to assert his rights by force of arms. 
Plato would take no part in the adventure, but the young 
hotbloods of the Academy were eager in the cause of their 
fellow-student, among them Plato's nephew, Speusippos, 
and Eudemos of Cyprus, the friend after whom Aristotle 
named his dialogue on immortality. 2 All preparations 
were completed by the summer of 357 B.C., but diffi- 
culties began at once. Herakleides, who had gone into 
exile after the incident described above, would not subor- 
dinate himself to Dion and remained behind. With only 
800 men Dion set sail for Sicily. Philistos was waiting 
for him in the Adriatic ; but Dion eluded him by sailing 
straight across the sea instead of following the usual coast 
route. Once landed in Sicily he received accessions of 
strength from every side. Dionysios, who had not ex- 
pected an attack in this direction, was in Italy, and 
Dion made himself master of Syracuse. All might now 
have been well had Dion been a little more concilia- 
tory. Herakleides arrived on the scene and had to be 

1 This may be why Dion had tried to secure the succession for the 
sons of Dionysios I. by Aristomache. They were much younger. 

2 Eudemos lost his life in one of the combats round Syracuse. 


given a share in the government, but this proved a 
constant source of weakness, and led at one time to the 
temporary deposition of Dion. This is not the place 
to recount the wretched details of the three-cornered 
struggle between Dionysios, Dion, and Herakleides ; it 
will be enough to indicate its result. Herakleides was 
murdered at the instigation of Dion, and Dion himself 
fell by the dagger of Kallippos, an Athenian and a 
member of the Academy, who had been his most con- 
fidential adviser. Kallippos only held power for about a 
year, when he was once more expelled by Dion's partisans. 

Plato felt deeply the discredit which the treachery of 
Kallippos had brought upon Athens and the Academy, 
but he never wavered in his belief in Dion's integrity. He 
was well aware of the defect in his character which has 
been pointed out, 1 but he continued to regard him as per- 
fectly sincere and disinterested in his political action. In 
support of this estimate it may be observed that it would 
have been comparatively easy for Dion, who was closely 
related to the royal house, to brush Dionysios aside at the 
beginning of his reign and seize the power for himself. 
Instead of that he did his best, in conjunction with Archytas, 
to fit the young prince for the position he was called upon 
to occupy. If he was embittered by the return he received 
for this act of self-abnegation, we can hardly wonder at it. 
His property had been confiscated, and his wife had been 
compelled to marry another man. 

223. The overthrow of Kallippos was the occasion of 
Plato's last endeavour to do something for Sicily. The 
partisans of Dion asked him for advice with regard to the 
settlement of the constitution, and this gave him the 
opportunity of writing the two open letters to which we 

1 In his letter congratulating Dion on his success (Epistle iv.) Plato 
tells him that some people think him too deficient in complaisance, and 
warns him against this fault (321 b). He is very anxious that the rule 
of Dion should do the Academy credit. He reminds him that the "you 
know whos " (rous olcrOa B^TTOV 320 c) are expected to surpass others 
even more than grown men surpass children. 


owe all our knowledge of these affairs. The first (Epistle 
vii.) is a dignified defence of his own political attitude 
throughout life, and it bears witness at once to his dis- 
appointment in men whom he had trusted, and to his 
unshaken confidence in his principles. He is willing to 
advise the partisans of Dion, if they are really sincere in 
their desire to realise Dion's plans. He clearly does not 
feel sure of them. In the second letter (Epistle viii.) he 
suggests, however, a scheme for the government of 
Syracuse, in which Dionysios himself was to be asked 
to take a share, if he would accept it, along with Hippa- 
rinos, his brother, and Hipparinos, the son of Dion. It 
need hardly be said that this proposal was too statesman- 
like to be accepted by embittered party men, and so the 
Syracusan Empire broke up for the time being. As Plato 
saw, it was in danger of falling into the hands of the 
Carthaginians or the Oscans. 1 

We have seen how very nearly Plato came to succeed- 
ing. At the very least he might have done for Dionysios 
what the Pythagorean Lysis did for Epameinondas. It 
was said at the time that the prosperity of Thebes at this 
date was due entirely to the philosophers. 2 And he might 
have done even more with more promising material. If 
it had been an Alexander of Macedon that Plato had to 
deal with instead of a Dionysios, a Greek king would have 
been ruling at Carthage before many years had passed. 
As it was, it was left for the Romans to carry out the task 
which seemed to fall naturally to the ruler of Syracuse, 8 and 

1 Ep. viii. 353 e. 

2 Alkidamas said : Q-fj/Bycriv a/xa 01 Trpocrroirai <j>t,X6a-o<f>oi lyevovro 
KOL eTjSat/Aov^orev y 7roXts (Ar. Rhet. i^tySb, 1 8). 

3 The First Punic War broke out just eighty years after the final 
expulsion of Dionysios II. from Syracuse by Timoleon. Plato did not 
live to see either the brief restoration of Dionysios (345 B.C.) or his final 
overthrow (344 B.C.). After that Dionysios lived the life of a dilettante 
at Corinth, where Aristoxenos saw him, and asked him the cause of his 
quarrel with Plato. Dionysios answered that no one tells a tyrant the 
truth, and that he had been robbed of Plato's goodwill by want of frank- 
ness in his so-called friends (Plutarch, Timokon^ 15). 


that brought about the division between Eastern and 
Western Europe which, to all appearance, will be the 
great political problem of the immediate future. 

The Laws. 

224. It must not be supposed, however, that Plato's 
attempt to make a constitutional ruler of Dionysios bore 
no fruit, even at the time. It was the immediate occasion 
of his undertaking his longest and most comprehensive 
work. It is true that a credible tradition represents the 
Laws as having been published after Plato's death by 
Philip of Opous, and it is likely enough that he never gave 
the finishing touch to the work. That is quite consistent, 
however, with its having been begun a good many years 
earlier. It is a treatise which goes into great detail, and 
which must have called for considerable study of existing 
codes of law. Now in Epistle iii. (316 a), written shortly 
after 360 B.C., we are told expressly that Plato had been 
working with Dionysios at the cc preambles" (Trpooi^ia) to 
laws during his second visit to Syracuse. This is explained 
by a passage in the Laws itself (722 dsqq.\ where we are 
told that the legislator ought always to preface his laws by 
a " prelude" (irpooifjuov) in which he explains their motive. 
That gives us some insight into Plato's method of teaching 
politics and jurisprudence, which is quite in accordance 
with the doctrine of the Statesman. In order to frame a 
code of laws on any subject, we must first of all lay down 
clearly the general principles which are to guide us, and 
then go on to embody these in detailed enactments. The 
general principles will as far as possible be such as would 
be approved by the ideal ruler who can dispense with laws 
altogether ; the particular enactments will take account of 
the circumstances of the state for which they are intended. 

The fiction of the dialogue is that a colony is to be 
established in Crete on a deserted site, and the magis- 
trate of Knossos who is charged with the duty of 
legislating for it is represented as consulting an Athenian 


Stranger and a Spartan on the subject. The very first 
questions asked before legislation in detail is attempted are 
whether the new city is on the coast or inland, whether the 
soil is fertile or not, and the like (704 a sqq^ There is no 
attempt to legislate for a city in the abstract ; we are dealing 
with a particular colony, and we have to take account of 
all the special circumstances affecting it. 

225. There is no work of Plato's which has been so 
little appreciated as the Laws, and yet it contains much of 
his maturest thought which we should otherwise know 
nothing about, and embodies the results of a long and 
varied experience of human life. It is, of course, im- 
possible to summarise it here ; all that can be done is to 
suggest certain points which may help the reader to a 
juster view of what Plato himself probably considered his 
most important work. 

He still believed, in spite of his disappointment with 
Dionysios, that the co-operation of a tyrant with a philo- 
sopher would result in the greatest blessings for the 
Hellenic nation, and he reasserts this conviction emphati- 
cally (709 e). Failing that, however, much might be 
hoped from the influence of philosophy on law-givers 
and framers of constitutions. He did not, therefore, 
think it an unworthy use of his last years to codify what 
seemed best to him in Greek Law, public and private, and 
especially in the Law of Athens, supplementing it with 
legislative proposals of his own. To understand this we 
must try to realise the condition of the Greek world at the 
time. We are not accustomed in this country to systematic 
legislation (what the Greeks called vojmoOecria)^ though such 
things as the Code Napoleon may give us a notion of what 
is meant, but it was very familiar to the Greeks. Every 
colony had a written constitution and a code of laws, and 
the task of framing these was regularly entrusted to a 
single individual or a small commission. The situation 
presupposed in the Laws was of almost everyday occur- 
rence, and there is nothing extravagant in the idea that a 
man like the Athenian Stranger who is more or less Plato 


himself should be able to give valuable assistance in such 
circumstances. It is certain, indeed, that many of the men 
who gave laws to the Greek States at this time were mem- 
bers of the Academy, and that several States applied to the 
Academy for an expert legislator when they were amend- 
ing their constitutions. 1 The purpose of the Laws is, 
therefore, an eminently practical one, and the work is 
designed to meet a real need of the time. 

226. No doubt it may seem strange to a modern reader 
that Plato should devote so much attention as he does 
to minute police regulations about water-supply and the 
picking of ripe fruits by the passing wayfarer. As to 
that, there are two remarks to be made. In the first place, 
one of Plato's most deeply rooted convictions is that all 
human affairs are very insignificant in comparison with 
the immensity of the world, and that the events of the 
day are only an incident in the history of mankind through 
countless ages. Sometimes he feels that Man is perhaps 
no more than a plaything of God, and that human life is 
not after all a serious thing. Unfortunately, whether it is 
serious or not, we have got to take it seriously (803 b), 
but it is absurd to suppose there is much to choose 
between one department of it and another in point of 
worth and dignity. Nothing is too humble, as nothing is 
too exalted, for the philosopher's attention. 

Closely connected with this is his belief that homely 
examples are often the best to illustrate important prin- 
ciples. He had learnt that from Sokrates, and he had 
discussed the matter in the Statesman. This is particu- 
larly the case in jurisprudence. Jurists, who presumably 
know their business, do not quarrel with the Institutes 
for their minute discussions of the ownership of stray 
animals and swarming bees. It is not to be supposed that 

Adv. Col. 11260 nXtxTWV & rtov ITCU/OCOV 4cwrrTtAev 
wci/ 'Apttrrtijru/AOV Sta/v-ocr/x^crovTa TT)V Troktreiav, 'HAetoi? 
Meve&ov & TLvaiois. E{5Soos 8e K^Stcns KOL 

<E>0/>/uo>va, Meve&^ov & TLvppaiois. E{5So ^ 

Srayct/nTats, HAartovos OI/TCS awr^ets, i/o^ous 


these questions were treated entirely for their own sake 
by the Roman lawyers ; it is because such simple instances 
are the best for the purpose of bringing out the funda- 
mental principles of law. 

This brings us to another very important point. We 
have seen that many of Plato's associates became lawgivers, 
and it is hardly too much to say that his work is the foun- 
dation of Hellenistic Law. That explains the fact, which 
was perfectly well known to some of the older jurists like 
Cujas, though it is often overlooked at the present day, 
that many features of Roman Law are derived from this 
source. 1 The direct influence of Greek philosophy on 
Roman Law has probably been overestimated, but its 
indirect influence has hardly been done justice to. The 
way in which this came about was as follows. When the 
Romans came into closer contact with non-Roman peoples, 
that is to say, especially with the Greek communities of 
Italy and Sicily, it was found that the principles of their 
civil law could not be applied easily to the relations between 
Romans and foreigners or to the relations of foreigners 
with one another. Hence arose the jus gentium^ which, in 
its origin, was a sort of common law of Italy. This was 
administered by the praetor peregrinus and embodied in his 
edict, which was simply an announcement of the principles 
on which he intended to decide certain cases. The edict 
was handed down from praetor to praetor with such modi- 
fications as were required from time to time, and ultimately 
became a regular body of law, the jus honorarium. It was 
inevitable that many of its provisions should be modelled on 
the laws of the Hellenic states with which the Romans came 
in contact, and these in turn were profoundly influenced 
by the jurisprudence of the Academy. Now that Hellen- 
istic law is becoming better known from the papyri, we may 
confidently anticipate some valuable discoveries in this field. 

1 See Cuiacii Comm. in lib. xlix. Pauli ad Edictum, ad ad Namusam 
et seq. : multa . . . auc tores nostri ex Platone mutuati sunt. Examples 
are given in Observationum lib. xxiv, c. 24, 



227. In the next chapter we shall be dealing with the 
most abstract aspect of Plato's philosophy, so it will be 
well to give here a brief sketch of the educational system 
recommended in the Laws. This will keep us in mind 
that these highly abstract speculations went hand in hand 
with the most intense interest in concrete detail. It will 
also be useful from another point of view. The educa- 
tional theories of Plato are chiefly known from the 
Republic, and it is often forgotten that there is a much 
fuller and more practical treatment of the subject in the 

The first thing to secure is that babies shall be straight 
(788 d), for everything depends on the start. A human 
being^may go on growing till he is twenty, but quite half 
of this growth is accomplished in the first five years. 
Now growth implies nourishment, and the nourishment 
of babies is very great in proportion to their size. It 
follows that they must have a great deal of bodily exercise 
up to the age of five. The simplest way of putting this 
is to say that babies should live as if they were always at 
sea. Even nurses know that from experience, for when 
they wish to put babies to sleep they employ action, not 
rest, for the purpose. They shake them up and down in 
their arms, and they do not use silence, but sing to 
them. The Korybantic purifications depend on the same 
principle (790 d). 

The next point to notice is that small babies scream 
and kick, while larger ones shout and jump about in 
a disorderly fashion. For three years babies can only 
express their wants by crying ; and as three years is 
a considerable portion of a human life to spend well or 
ill, education must start from this fact, and build upon it. 
Pleasure and pain are the only feelings young children 
know, and we might suppose it the right thing to give 
them all the pleasure and save them all the pain we can. 
That, however, is wrong. What we wish to train them 


to is that state of calm which is as far removed from 
positive pleasure as from pain. In order to do this we 
must take advantage of the fact that from the very 
earliest age children take pleasure in tune and time. 
These two things must therefore be our chief educational 
instrument for the first three years of life ; for, by 
developing this instinct, we can gradually transform the 
natural screams and shouts into song, and the kicks and 
jumps into dance. Punishment should begin at the age 
of three, but we must be careful not to employ forms of 
punishment which will produce anger and sullenness. As 
to games, they are instinctive at that age, and when a few 
small children are brought together, they will invent them 
of their own accord. It is best to leave them to do so. 

From three to six children should be taken to the 
religious services of their village, and this at once raises 
the thorny problem of nurses. There must be a com- 
mittee of twelve ladies appointed by the head of the 
Education Department to supervise all the nurses. They 
will divide the country into districts, and each will visit 
all the temples and celebrations in her own district, at 
least once a year, to see that the nurses behave. It is a 
good plan for the grandparent to live at some distance 
and have the children sent to visit them. In that way it 
is possible to make sure that they really do get the outing 
they are supposed to get. 

The education of boys and girls should be separate 
from the age of six, for at that age they begin actual 
lessons. The boys are to be taught riding and archery 
and the use of the sling. The girls are also to be taught 
the use of arms as far as possible. We must also get rid 
of the superstition of mothers and nurses that the right 
hand is to be preferred to the left. It makes us only half 

The chief instruments of education at this stage will be 
music and gymnastics, for which we have prepared the 
children by the use of time and tune and by shaking them 
when they were small. Gymnastics has two main divisions, 


dancing and wrestling. Music has two functions one 
the accompaniment of the noble words of the poets, the 
other the accompaniment of dances and other exercises 
of the limbs. We must not teach the children anything 
elaborate or professional, but only simple physical drill 
with simple songs, taking as our model what is required 
in war and the service of the gods. The question of 
games and toys becomes more important at this age. 
The main thing is that each generation should play the 
same games and have the same toys as the last, for only 
so can the spirit of the constitution be preserved. The 
greatest of all revolutionaries is the man who invents new 
games and finer toys, for the boy who has played different 
games in youth will grow up a different sort of man. 
In things which are not in themselves bad change is 
dangerous, and therefore the preservation of the old 
games is a fundamental interest of the state. As to 
music, we must take it as our guiding principle that 
rhythms and melodies are imitations of character. They 
are the most direct imitation there is of anything far 
more direct than painting and sculpture, for instance 
but what they imitate is not the outward appearance 
but disposition of soul. These, then, must be preserved 
unaltered too. New melodies and rhythms will destroy 
the spirit of the constitution. Tragedy will be ex- 
cluded, of course. We cannot allow competing choruses 
to blaspheme in the immediate neighbourhood of the 

The difficult task of selecting songs and dances will 
be left to a jury consisting of men over fifty, who will 
accept or reject the old ones, or, if necessary, call in 
expert assistance to correct their melody and rhythm. If 
the children are once accustomed to the sober and 
ordered Muse, when they hear the opposite kind of 
music, the sweet kind, they will think it only fit for 
slaves. On the other hand, if they have been habituated 
to the sweet Muse in early life, they will find true music 
cold and harsh. There must be separate songs for boys 


and girls, differing in pitch and time. The boys' music 
will imitate the proud and brave character, the girls' the 
modest and pure. Gymnastics must be taught to girls 
also. There is no reason for supposing that riding and 
gymnastics are suitable for boys and not for girls. It is 
true that women are not so strong as men, but that is 
no reason for their not being made to do what they can. 
A state that makes no call upon its women for military 
service is not much more than half as strong as it might 
be made at the same expense. It would be better that 
they should be relieved to some extent from household 
occupations, which might be simplified by the introduc- 
tion of co-operative methods. At any rate, the human 
race should be freed from the disgrace of being the only 
one in which the females are incapable of defending the 
life of their young. 

We have not yet touched on the manner in which these 
things are to be taught. It is not merely a technical one. 
Everything depends on the object we have in view. 
Just as a shipbuilder constructs a ship with a view to a 
certain kind of voyage, so our educational methods must 
be determined by a view of the best way to make the 
voyage of life. Perhaps it does not matter from the point 
of view of God, but we must at least play the game if it 
is one, and who knows but it may be more. Even if 
men and women are God's playthings, that is, after all, 
the best thing about them. The trouble is that people 
draw the distinction between jest and earnest, work and 
play wrongly. They suppose, for instance, that war is 
earnest and peace is not. That is wrong. Peace is more 
earnest than war, and a great deal that is taken for play is 
really the highest kind of work. 

The question of school buildings is of great importance. 
The teachers must have salaries, and therefore (this is 
very Greek) they must be foreigners. Education must 
be compulsory. It cannot be left to the fathers of 
families to educate their children or not as they please, 
for they belong even more to the state than to their 


fathers. So far we have been dealing with what 
we should call elementary education, which was all the 
education most men had in Plato's time. 

228. But now comes the question what our young people 
are to do now that their preliminary training is finished. 
Is there something further, or are they to live the life of 
cattle being fattened for the market ? Certainly not. Now 
is the time for real hard work ; all the rest, including the 
military training, has really been play. There is no time 
to lose. In very truth every day and night of our lives, 
if devoted to that alone, is barely sufficient for a complete, 
or even an adequate education. The employment of each 
day must therefore be carefully ordered from one sunrise 
to the next. It would be unseemly for the legislator to 
enter into domestic details, but we may say at once that 
it is monstrous for those who are to guard a city to sleep 
all night, and that it is not proper for the mistress of a 
house to be wakened by her maids. She should be up first 
and see that the maids are up. A man who is asleep is 
worthless, and he who cares most to be alive and thinking 
keeps awake longest. It is wonderful how little sleep we 
need when we get into the habit of doing with little. The 
boy must therefore go to school before sunrise. He wants 
careful watching ; for he is the most awkward of beasts to 
handle. That is just because he has what other beasts 
have not, a native spring of thought in him which is not 
yet settled or clear. Boys will now study things written, 
and not all of them in metre. Along with that will go at first 
the tuning of the lyre (not necessarily the playing of it), so 
much reckoning as is useful for war and housekeeping, and 
a certain amount of astronomy, enough to make the 
calendar intelligible. These things are not to be confused 
with the sciences, which come later. 

^- The question arises how far a man who is to be a good 
citizen must go in these subjects, A boy should begin 
reading and writing at the age of ten and spend three years 
on them ; music need not be begun till he is thirteen, and 
should be continued for three years. These times should 

3 io THE LAWS 

be made compulsory whether the boy or his father has any 
taste for the subjects or not It will be enough if the boys 
can read and write intelligibly; it is only in cases of special 
talent that we should encourage a higher degree of excel- 
lence. The time and trouble it takes are better spared for 
the higher studies. 

That the boys will read poetry of the right sort is a 
matter of course, but prose seems a very dangerous thing. 
Even as to poetry there is the question whether it should 
be read in masses and whole poets learnt by heart, or 
whether we should use books of extracts and make our 
pupils commit these to memory. But, as has been sug- 
gested, the real difficulty is the educational use of prose. 
Books about the principles of legislation may certainly be 
read, but the works of philosophers and scientific men are 
not safe at this stage. All these things will be regulated by 
the head of the Education Department, but he will have 
expert advice on technical questions. He will not allow 
the experts to dictate to him on general principles, but will 
consult them as to the methods of carrying them out. 

229. We come now to the higher studies, beginning 
with Mathematics, in its three chief divisions of Arithmetic, 
Geometry, and Astronomy. Only a small number will 
pursue these studies to the end, those, namely, who show 
themselves fit to become members of the Nocturnal Council, 
but the prevailing ignorance of them can only be described 
as " swinish" (819 d). And that is not the worst. Most 
teachers treat mathematical subjects in the most perverse 
manner, and the greatest evil is not total ignorance, but 
much learning and knowledge misdirected. Most people 
take it for granted that all lengths, breadths and depths 
are commensurable, whereas it is really the problem of 
incommensurability that should hold the first place in 
mathematical education. The study of questions arising 
out of this is a far better game than backgammon. The 
teaching of astronomy must be reformed on similar lines. 

We may easily miss the significance of Plato's proposals 
as to the education of boys and girls from the age of ten 


onwards. We must remember that in his day there were 
no regular schools for young people of that age. They 
were taken to one teacher for music-lessons and to another 
to be taught Homer, and there was no idea of coordinating 
all these things in a single building under a single direction 
with a regular staff of teachers. By founding the Academy 
Plato had invented the university, and now he has invented 
the secondary school. In consequence we find such schools 
everywhere in the Hellenistic period, and the Romans 
adopted it with other things, quaintly translating the Greek 
term o^oA?? by Indus. That is the origin of the medieval 
grammar school and of all that has come out of it since. 
It will be seen that the Laws is not a work we can afford 
to despise if we wish to understand Plato's influence, but 
it is time to turn to a very different side of his activity. 



230, It is by no means easy for us at the present day 
to interpret the central doctrine of Plato's philosophy. As 
we have seen ( 162), he did not choose to commit it to 
writing, and we are almost entirely dependent on what 
Aristotle tells us. What makes matters worse is that 
Aristotle is a very unsympathetic critic of Plato's teaching, 
and that he looks at it too much in the light of certain results 
to which it had led in the Academy of his own day. In 
one place he complains that the men of his time (ol vuv) 
had replaced philosophy by mathematics. 1 That was re- 
pugnant to him as a biologist, and he made the teaching 
of Plato responsible for it. We shall have to see how far 
he was justified. 

In dealing with Aristotle's evidence, it is necessary to 
make two distinctions. We must, in the first instance at 
least, distinguish (i) between doctrines attributed to Plato 
by name and doctrines vaguely stated to be those of 
" some," a way of speaking which may include Pytha- 
goreans and the contemporary Academy. We must also 
distinguish even more carefully (2) between statements as 
to facts which must have been well within Aristotle's 
knowledge and his interpretation of these facts. When he 
tells us, for instance, that Plato held numbers to be unadd- 
ible, we are bound to believe him. He could not have 
made such a statement unless it was true and was known 

l Met. A, 9, 992 a, 32 : yeyo/ rot /xa^/xara rots vvv vj faXocrofoa. 


to be true by his contemporaries. On the other hand, 
when he tells us what Plato really meant by this, we have 
to remember that he is one of those people who always 
know what another man means better than he knows him- 
self. Above all, when he describes the historical origin of 
any doctrine, we must bear in mind that he is speaking of 
things he could know nothing about except from inference 
or hearsay. These obvious distinctions are often ignored. 
Speculations as to the influence exercised on Plato by 
Sokrates and Kratylos years before Aristotle was born are 
quoted as evidence of fact, and at the same time a philo- 
sophy is expounded as Plato's, which differs in the most 
important points from that which Aristotle says he heard 
from his own lips. 

One thing, at any rate, seems clear. Aristotle knows 
of but one Platonic philosophy, that which identified the 
forms with numbers. He never indicates that this system 
had taken the place of an earlier Platonism in which the 
forms were not identified with numbers, or that he knew 
of any change or modification introduced into his philo- 
sophy by Plato in his old age. 1 That is only a modern 
speculation, Aristotle had been a member of the Academy 
for the last twenty years of Plato's life, and nothing of 
the kind could have taken place without his knowledge. 
We may be sure too that, if he had known of any such 
change, he would have told us. It is not his way to cover 
up what he regards as inconsistencies in his master's teach- 
ing. If the *' theory of Numbers" had been no more than 
a senile aberration (which appears to be the current view), 
that is just the sort of thing Aristotle would have delighted 
to point out. As it is, his evidence shows that Plato held 
this theory from his sixtieth year at least, and probably 

1 In M. 4. 1078 b, 9 sqg., it seems to me impossible to identify those 
who "first said there were forms" with Plato, though it must be admitted 
that things are said of them which are said of Plato in A. 6. The ex- 
planation is, I think, that in both cases Aristotle is thinking primarily of 
the tlS&v <tAoi in the Phaedo (cf. p. 280). 


231. It is certain, then, that Plato identified forms 
and numbers ; but, when we ask what he meant by this, 
we get into difficulties at once. In the last two books of 
the Metaphysics (M and N), which deal expressly with 
the objects of mathematics (ra /xa^/xar^a) and with forms 
and numbers, the name of Plato is only mentioned once 
(1083 a J 33)5 an ^ the doctrine there attributed to him is 
that numbers "are not addible to one another" (ov o-vjm/3\y- 
TOW elvcu TOV$ apiOjuovs 7rpo$ a\\ij\ovi). In an earlier passage 
(1080 a, 12 sqq^ three versions of the doctrine that 
numbers are "separate" (yfopurTo) and the first causes of 
things are given as the only possible ones, but no names 
are mentioned. We are even told (1081 a, 35) that one 
of these versions had never been held by anybody, which 
does not prevent Aristotle (if he is the author of these 
books) from refuting it as vigorously as the other two. 
Obviously we cannot make anything of this for the present, 
and it is unsafe, at least in the first instance, to use these 
books as evidence except for the single doctrine attributed 
in them to Plato by name. 

232. There is, however, a chapter in the First Book 
of the Metaphysics (A. 6) which seems more hopeful. It 
is the only place where Aristotle professes to give a careful 
statement of Plato's philosophy, attributing it to him by 
name and distinguishing it from other systems. The 
method he adopts is to compare Platonism with Pytha- 
goreanism, which, he says, it followed in most respects 
(ra TroXXa), though it had two peculiarities (Ktox UXdrcwoi) 
which distinguished it from " the Italic philosophy." 
These two points of difference were as follows : (i) The 
Pythagoreans said that numbers were things, while Plato 
held not only that sensible things were distinct from 
(jrapa) numbers, but also regarded the objects of mathe- 
matics as distinct from both and intermediate between 
them. (2) The Pythagoreans held the matter of numbers 
to be the Unlimited and their form the Limit ; Plato 
regarded the elements of number as the One and the dyad 
of the Great-and-Small. 


These two points are all that Aristotle regards as really 
peculiar to Plato ; for he looks upon the substitution of 
the term " participation " for " imitation " as a merely 
verbal difference. Both the Pythagoreans and Plato left 
it an open question (atpeicrav ev KOLVW fyreiv) what imitation 
or participation of things in forms could be. That is the 
outline of the chapter, but it is somewhat confused by a 
long parenthesis intended to show that the first difference 
between Plato and the Pythagoreans was due to the influ- 
ence of Herakleitos (through Kratylos) and Sokrates. 
That may or may not be correct, but Aristotle's statements 
on this subject do not stand on the same level as his account 
of the peculiarities themselves, which he must have heard 
Plato expound. 

I. Forms y Mathematical and Sensibles. 

233. The first of these peculiarities is, then, that, while 
the Pythagoreans said numbers were things, Plato regarded 
sensible things as distinct from numbers, and made the 
objects of mathematics intermediate between the two. It 
is important to observe that Aristotle is here contrasting 
Plato with the Pythagoreans and not with Sokrates, who 
is only introduced to explain his divergence from the 
Pythagorean theory of numbers. It is also to be noted 
that by " Sokrates " Aristotle means, as he usually does, 
the Sokrates of the Phaedo. We are expressly told 
(987 b, 29) that the distinction made between numbers 
and the sensibles and the " introduction " (etVaycoytf) of 

the forms was due to the practice of " considering things 
in statements" (Sta TYJV ev TO?? Ao'ycu? eyevero cr/te-vj/w) and 
that is as clear a reference as can be to the new method 
introduced by Sokrates in that dialogue (99 e sqq.}. We 
are also told that the predecessors of Sokrates were un- 
versed in dialectic, and that is explained by what has been 
said above (987 a, 20) about the Pythagoreans. They 
began, we are told, to discuss the "What is it ? " of things 
(TO rl co-rev ;), and to define them, but in a naive and 


superficial way. Sokrates introduced universal definitions 
and busied himself with ethical matters instead of with 
nature as a whole, and it was Plato's acceptance of his 
method that made it impossible for him to follow the 
Pythagoreans in identifying numbers with things. He 
had convinced himself of the Herakleitean doctrine that 
sensible things were in flux, and he saw that the definitions of 
Sokrates could not apply to them, so he gave the name of 
forms to something other than sensible things, and said 
that sensible things were distinct from these (vrapa raura) 
and were called after them ; for the multitude of things 
sharing the same name as the forms were what they were 
in virtue of their participation in these forms. It will be 
observed that in this passage Aristotle insists rather on the 
distinction of sensible things from the forms than on that 
of the forms from sensible things, and he implies that this 
is what distinguished Plato from Sokrates. We have seen 
reason already for believing that Sokrates recognised no 
reality in sensible things apart from the forms, and Aris- 
totle's language here confirms this view. Of course it is 
equally true to say, as Aristotle usually does, that the forms 
are distinct from the sensible things, but it is significant 
that, when he first has occasion to mention the point, 
he emphasises the other side of the distinction. 

234. Closely connected with this separation (xf^fttrjuiog) 
of sensible things is what Aristotle calls the "introduction" 
(elcrayuyfy of the forms. This term does not imply that 
Plato invented them. The metaphor is, I believe, derived 
from the use of the word for bringing on the stage or 
cc producing," and the suggestion appears to be that the 
ethical inquiries of Sokrates had made it necessary to 
assume certain universals which were not numbers, and 
these, of course, would be separate from the things of 
sense just as the numbers were. The Pythagoreans had 
defined Justice, for instance, as a square number, but 
Sokrates had shown that we must postulate a special form 
of Justice (avro o eo-n Slxaiov). That is not mentioned as 
an innovation of Plato's. The only difference which is 


implied between Sokrates and Plato is that the latter 
separated sensible things from the forms while the former 
did not. That is stated in so many words in the Tenth 
Book (1078 b, 17), though it is also said (1086 b, 3) that 
Sokrates gave the impulse to (e/c:V^cre) this separation. He is 
commended for not going further, and it is implied that 
his doctrine was much the same as Aristotle's own. That 
can hardly be historical, but Aristotle may have thought it 
a legitimate interpretation of the second part of the Phaedo^ 
where the forms are certainly in things. It seems to me 
a far more serious anachronism to represent Sokrates as 
seeking for universals (ra /caQo'Aoi/), a term not yet invented, 
than to represent him as seeking for " forms." It is worse 
still to make him talk about "concepts." 1 Realism is prior 
to Conceptualism, and I doubt very much whether anyone 
ever " hypostatised concepts." As we have seen ( 195), 
Conceptualism is tentatively put forward in the Parmenides 
as a solution of the problem of participation, but it is 
rejected at once. 

235. This parenthesis, then, is at best Aristotle's 
speculative reconstruction of history from his own point 
of view, and throws very little light on his definite state- 
ment that Plato not only made numbers distinct from 
sensible things, but also made the objects of mathematics 
intermediate between them. It is that statement of Aris- 
totle, and not his historical notes upon it, which we have 
really to interpret. He tells us further that the objects 
of mathematics differed from the things of sense in being 
eternal and immovable and from the forms in being many, 
whereas each form is one and unique (aim? &/ f^ovov). If 
we can interpret that, we shall know what Plato's "separa- 
tism " (x<jopicrju<oi) really meant. 

The difference between the objects of sense and the 
objects of mathematics is a simple matter, and is fully 
dealt with in the Phaedo. The mathematician is not really 
speaking about the sensible diagram he traces in the sand. 

1 The term A<5yo$ cannot possibly mean " concept." So far as there is 
any Greek word for " concept " at this date, it is v6r)pa. 


The sensible circle is only a rough " image " (elSooXov) of 
what he really means. In the Phaedo, however, the 
objects of mathematics are certainly regarded as forms, 
and we have now to ask what is meant by distinguishing 
them from the forms. It cannot, of course, be meant 
that mathematical forms are on a lower level than others. 
That is the last thing Plato would think of, and the point 
is rather that they are on a higher level. The object of 
the mathematician's reasoning is not, indeed, the sensible 
circle, but neither is it the circle, the form of circularity. 
He speaks of circles of greater or smaller radius, and even 
of two circles intersecting one another. Mathematical 
reasoning, then, has to do with many circles, whereas the 
circle is one and one only. In the same way, the triangle 
about which we reason is either equilateral, isosceles or 
scalene, but the triangle is none of these. In fact, it is 
really the circles, triangles, etc., of which the geometer 
speaks that are the " many" which partake in the forms. 1 
And this is even truer of numbers than of figures, the 
spatial character of which has something of the sensible 
about it. We speak of adding two and two to make four, 
as if there were many twos. It is clear that we do not 
mean by these twos the pebbles or counters we may use 
to symbolise them, but neither do we mean the number 
two. There is only one number two, the form of two or 
the dyad. The arithmetician's twos, however, are even 
less like things of sense than the geometer's circles ; they 
are the nearest approach we can get to the purely 
intelligible. From this point of view, Plato's separatism 
is a good deal less arbitrary than Aristotle seems to 

236. This distinction, moreover, furnishes the real 
explanation of the doctrine Aristotle attributes to Plato 

1 There is a hint, perhaps unconscious, of this doctrine in the Phacdo, 
where Sokrates speaks of avra rot ra (74 c). These are not identical 
with the more or less equal things of sense nor yet with awo r& &rov. 
Probably such things as the two angles at the base of an isosceles triangle 
are meant 


by name, that numbers are " unaddible " 
When we say " two and two is four/' we mean that two 
units of a given kind added to two units of the same kind 
are equal to four units of that kind ; we do not mean that 
the number two added to the number two is the number 
four. That would be nonsense ; for the number two does 
not consist of two units nor does the number four consist 
of four units. Each number is a universal, and every 
universal is one and unique. The units we call "two" 
somehow partake in the number two, but it is not identical 
with them. There is only one number two. From this 
it follows further that the relation between the numbers 
themselves is not one that can be expressed by any additive 
formula. The number five is not the number four plus a 
unit. The relation of four and five is simply one of 
priority and posteriority. What, then, are the " two and 
two" which we say make four ? The answer will appear 
if we remember that the particulars of the mathematical 
sciences are objects of thought just as much as the 
universals. We can think particular "twos" without 
regarding them as inhering in any sensible substratum, so 
that the "two and two" which "make four" are dis- 
tinguished on the one hand from the "two and two 
pebbles " which make four pebbles, and on the other from 
the unique universal, the number two. 

It is clear, then, that numbers are unique forms, and we 
have seen some reason for thinking that they are forms in 
a pre-eminent sense. That is certainly the doctrine Aris- 
totle attributes to Plato, but we cannot understand it com- 
pletely till we have discussed the relation of the forms of 
number to the other forms. That brings us to what 
Aristotle regards as the second peculiarity (ISiov) of Plato's 

1 1 am much indebted here to Professor Cook Wilson's article in the 
Classical Review, vol. xviii. (1904) pp. 247 sqq. 


II. The One and the Indeterminate Dyad. 

237. The Pythagoreans had regarded the Limit 
and the Unlimited (airetpov) or Continuous as the elements 
of number, and therefore as the elements of things. Plato 
substituted for these the One and the dyad of the Great- 
and-Small. The only difference, according to Aristotle, is 
that the Pythagorean Unlimited was single, whereas Plato 
regarded the "matter" of numbers, and therefore of things, 
as dual in character. It also follows, as Aristotle points 
out elsewhere, from Plato's separation of numbers and 
things that there will be what he calls "matter" in the 
numbers as well as in things. This is called the Inde- 
terminate dyad (ao/wro? Svdi) l to distinguish it from the 
Determinate dyad, which is the number two. From this 
dyad the numbers are generated as from a sort of matrix 

238. Now it is at least clear that the term Indeterminate 
Dyad is a new name for Continuity, and it expresses more 
clearly than the old term Unlimited its twofold nature. 
It not only admits of infinite "increase" (autyi), but 
also of infinite "diminution" (m(9a/poV). s That is why 
it is also called the Great-and-Small. The new idea which 
Plato intended to express was that of the infinitesimal, 
the infiniment petit. The introduction of this conception 

1 The use of this term is not attributed to Plato by name, but Met. 
T 09 1 a, 4 seems to imply that he used it. 

2 Aristotle's account of the way in which the numbers are generated is 
extremely obscure. Mr. George A. Johnston has suggested a most 
interesting explanation of the matter, which. I have his permission to 
quote. We have seen (p. 53, n. i) that the ratio between the sides 
of successive oblong numbers (Le. the sums of the series of even numbers) 
is always changing. It is a dyad, because it is always a ratio between 
two numbers ; it is indefinite because the ratio is always changing. The 
one, on the other hand, is the square root of the successive oblong 
numbers, ^2, J6, *Jiz, etc,, which are means between the sides of 2 
(2.: I), 6(3:2), 12 (4: 3), etc. 

8 Not necessarily by division (Sicupco-is). The term /ca<9cu/)e<w is 
more general, and covers subtraction (a<cu'/oeori$). It is used in the 
extract from Hermodoros given below, p. 330. 


involves an entirely new view of number. That need not 
surprise us; for we have learnt from the Republic that it is 
the business of Dialectic to "destroy the hypotheses " of 
the special sciences, and also that the hypothesis of Arith- 
metic is the series of natural integers, each consisting of so 
many equal and indivisible units, and each either odd or 
even. From our present point of view, these units and 
their sums belong to the "intermediate" region. They 
are not sensible, indeed, but neither are they numbers in 
the true sense. The destruction of this hypothesis allows 
us to extend the conception of number so as to include 
quantities which are not a sum of units (jAovdSwv 7rA??(9o?), 
and which are neither odd nor even. We have seen that 
it was the study of incommensurables that made this 
extension necessary. That is indicated by the prominence 
given to the study of quadratic surds in the Theaetetus. If 
cc irrationals" are once regarded as numbers, the old hypo- 
thesis of Arithmetic is destroyed. 

This is not, as I understand it, tantamount to making 
the numerical series itself continuous; for in that case 
number would be identified with the mere potentiality of 
plus and minus, which is the Indeterminate dyad. It does, 
however, get rid of the indivisible unit, which was the 
source of all the trouble about irrational numbers. We 
may now regard the origin of the numerical series, not as 
I but as o, and there is no reason for refusing to call such 
quantities as </2 and ^5 numbers. The best proof that 
this was really the step which Plato took is that Aristotle 
always insists against him that there is no number but 
number made up of units (povaSucds apid^oi). It follows 
that Plato maintained there was. 

239. The hypotheses of Geometry were, of course, 
submitted to a precisely similar criticism. The new view 
of number had really broken down the barrier which 
Zeno had erected between Arithmetic and Geometry, and 
the old view of the point as "a unit having position" 
(/uLovas Oeo-w e^oi/a-a) was superseded. Aristotle has pre- 
served a very important piece of information as to Plato's 


oral teaching on this subject. He tells us that Plato 
objected altogether to the conception of a point as being 
a mere "geometrical dogma," and preferred to speak of 
"the origin of a line" (<*px*i 7/xw 1 ??)' 1 That implies the 
view that the line is generated from the point by what we 
know from other sources was called " fluxion " (/5wro). s 
This corresponds to the doctrine that the numerical series 
has zero, not the unit, for its origin* In the same way, 
the plane is a fluxion of the line and the solid of the 
plane. On the other hand, Aristotle adds, Plato often 
postulated indivisible lines. 8 Aristotle says it is easy to 
refute this doctrine, and the later commentators throw no 
light upon it. No doubt the term is paradoxical, but 
not more so than " infinitesimals.'' What Plato meant 
was clearly that, if you postulate indivisible units and 
regard i as the origin of the numerical series, you are 
also committed to indivisible or infinitesimal lines as the 
spatial unit. All this brings us very close to Newton 
and Leibniz, and the historical connexion can still be 
traced. 4 

240. When we look at geometry in this way, we 
see that its spatial character tends to become irrelevant. 
It becomes a form of Arithmetic, dealing with continuity 
in general, whether spatial or not. This view is fully 
developed in the Epinomis, where we are told (990 d) 
that Geometry (which is said in passing to be " a very 
absurd name ") is really " an assimilation by reference to 

1 Met. A. 992 a, I : TOVT<) P.GV o$v r<p yevet (sc. r<p rwv crrty/xtov) KCU 

2 Slmpl. in Phys. p, 722, 28 (Diels) : r) ypafj.^ pv<rt$ 
Proclus in EucL i. p. 97, 6 (Friedlein). 

8 Met. ib. : rovro Se TroXAa/as mfe ras ard/xof s y/oa^/xas. 

4 The recently discovered Discourse on Method by Archimedes has 
thrown unexpected light on the development of the method of 
infinitesimals among the Greeks, See Milhaud, Nouvelles faudes, pp. 134 
sqq. 9 and especially p, 154. Cavalieri's "method of indivisibles" is 
the connecting link between Greek and modern higher Mathematics. 
Newton and Leibniz got their knowledge of the former from Wallis 
and Barrow. Wallis translates /fams by Jluxus. 


surfaces of numbers not similar to one another by 
nature." That is just the development of what we 
read in the Tkeaetetus (148 a), to the effect that certain 
numbers are incommensurable " in length " (w'/ce*), but 
commensurable " by means of the surfaces of which they 
are roots " (TO?? eTriTreSois & SvvavTat). In precisely the 
same way Stereometry is said to be the art by which 
certain numbers not naturally similar can be assimilated 
by being raised to the third power, Aristotle strongly 
objects to what he regards as the confusion of Geometry 
with Arithmetic. He insists that the proper hypotheses 
of each science must be left undisturbed, and that it 
is illegitimate to prove a geometrical proposition by 
Arithmetic. We may infer that Plato held otherwise. 

There is also a fragment of Plato's friend Archytas 
which puts the matter very clearly, and proves this was 
really the direction mathematical thought was taking at 
the time. He says (fr, 4) : 

I think that in respect of wisdom Arithmetic surpasses all 
the other arts, and especially Geometry, seeing it can treat 
the objects it wishes to study in a far clearer way. . . . 
Where Geometry fails, Arithmetic completes its demonstra- 
tions in the same way, even with regard to figures, if there is 
such a thing as the study of figures. 1 

241. In the last resort, then, geometrical figures are 
reduced to numbers, and these in turn are generated from 
the One and the Indeterminate dyad. What is new here 
is the assumption of a material element even in the forms, 
though that element is nothing more than abstract con- 
tinuity. The importance of this is that it tends to make 
the intelligible forms less disparate from the things of 
sense. It will be observed that it is precisely because 
Plato "separated" numbers from sensibles that it became 

1 Diels, VorsP i, p. 337, 6 /cat BQKGL a Aoyto-riKa Trorl rav <ro<tav 
TCOJ/ IJLGV aAAaV rc)(V(ov Kal TroXv 8ta<jbe/)iv, ara/o /cat ras yeco/^erpi/cas 
lrapy(rTepct) Trpay/Aareuecr&xt $ $eXt . . . /cat a 7rtA.t7rei ax a ycco- 
yuerpca, KOL a7ro5etias a Aoyt0"ri/ca tTnreXti /cat 6/^ws, cl /*ev 
rea TTpay^areta, /cat TTC/H rots et'Secrtv. 


possible for him to justify the world of appearance. This 
cannot be fully explained till the next chapter ; all we 
have to note at present is that the One combines with the 
Indeterminate dyad to generate the numbers, just as the 
forms combine with the Great-and-Small to generate 
sensible things. In that sense the elements of numbers 
were the elements of things. That is how Aristotle states 
it, and by great good fortune we possess a dialogue which 
must have been written while he was a member of the 
Academy, and which, though it deals primarily with 
another subject, and avoids the doctrine of form-numbers 
altogether, contains nevertheless some indications of Plato's 
thought at the time. I refer to the Philebus, one of his 
maturest works. 

The Philebus. 

242. From certain discussions in Aristotle's Ethics we 
get a hint of how the Philebus probably came to be written. 
Eudoxos had introduced into the Academy the heresy that 
Pleasure is the Good, a doctrine he probably received 
from the school of Demokritos, as Epicurus did at a later 
date. This raised considerable discussion, as was natural, 
and Speusippos in particular opposed Eudoxos vehemently, 
going so far as to maintain that Pleasure was an evil. 
Plato was interested, of course, and he did what he had 
not done for years ; he wrote a Sokratic dialogue on the 
subject. It was quite an appropriate theme for Sokrates 
to discuss, and there is little in the greater part of the 
dialogue which the Sokrates of the Gorgias or the Phaedo 
might not have said. On the other hand, Plato's dramatic 
power is no longer what it was, and the characteristic 
touches of the Sokratic manner are fewer than in the 
earlier dialogues, though more than is often supposed. 
Undeniably, too, the voice is sometimes that of the 
Stranger from Elea and sometimes that of the Athenian 
Stranger in the Laws, and in those cases we are justified 
in thinking that we have a hint at least of Plato's personal 


thought. I propose, for the present, to summarise only 
that portion of the dialogue which bears directly on the 
subject we are now discussing ; the general theory of 
Pleasure, though of the highest importance in itself, can 
only be adequately treated in connexion with the views of 
Eudoxos and Speusippos and of Aristotle's criticism of 
these. We get the impression from the Philebus that we 
are dealing with a dispute between the younger members 
of the Academy, in which Plato condescends to take part, 
though, by transferring the conversation to the fifth 
century and by making Sokrates the chief speaker, he 
avoids committing himself too much. 

243. Before the opening of the dialogue, Sokrates and 
Philebos (a youth of whom nothing is known) have been 
discussing the Good. Philebos has stated the position 
that the Good is Pleasure (^o^), while Sokrates has 
identified it with Thought (ippovyarii) or Wisdom. Philebos 
declines to argue the question, and Protarchos (another 
young man of whom nothing is known 1 ) undertakes to 
replace him as the advocate of Pleasure. It is not a little 
remarkable that the dialogue should be called after a per- 
sonage who takes practically no part in it. 

The two positions are more distinctly stated thus. That 
of Philebos is that Pleasure, understood in its widest sense 
as including joy, delight, and so forth, is the highest good 
for all living beings without exception. 2 That of Sokrates 
is that Thought, understood in its widest sense as in- 
cluding memory, right belief, true reasoning, and so forth, 
is the highest good for all living beings that are capable 
of it. The two positions agree in this, that both make 
Happiness (eu5a/*owa) a habit (?#) or disposition 

1 He is addressed as " son of Kallias " (19 b), but there is no ground 
for identifying him with one of the two sons of Kallias son of Hipponikos, 
mentioned in the Apology (20 b) as pupils of Euenos in 399 B.C. 

2 This seems to refer to the argument of Eudoxos that Pleasure must 
be the Good, since all things, rational and irrational, aim at it (Arist. 
EtL Nte. 1172 b, 9 $qq)* 


of soul. 1 It is further pointed out that there may prove 
to be a third habit of soul which is better than either 
Pleasure or Thought, in which case we must give the pre- 
ference to whichever of these two is most nearly akin to 
it (u a 12 a). 

244. Sokrates begins by calling attention to the fact 
that pleasures may be very unlike and indeed opposite, so 
that we cannot apply the same predicate to all of them, 
but it soon appears that it will be necessary to go deeper 
than this. We cannot, in fact, make any advance without 
coming to an understanding on the troublesome old ques- 
tion of the One and the Many. By this we do not mean 
the puzzle about the predication of opposite attributes like 
great and small, heavy and light, of the same subjects. 
That is child's play, and the solution has long been public 
property. Nor do we mean the question arising from the 
fact that every sensible thing has parts, and is therefore 
both one and many. The real difficulty is with regard to 
such units (monads^ henads) as horse, ox, beautiful, good 
(i.e. the a forms " of the Phaedo and the Republic). With 
regard to these we have to ask (i) in what sense we are to 
hold that each of these units really is, (2) in what sense 
we are to hold that each of them being one, and admitting 
neither coming into being nor ceasing to be, nevertheless 
is that one, 2 (3) in what sense we are to hold that these 
units can be present in the innumerable things of the 
sensible world, whether (a) in part, or () as wholes, so 
that (what seems quite impossible) they should be identical 
both in their unity and in their plurality (12 c 15 c). 

1 The terms ets and Sta#eor&s are taken from medicine. A "habit" 
is a more lasting "disposition" (Arist. Caf. 9 a, 8). The doctrine that 
Happiness is a habit of soul is characteristic of the Academy ; Aristotle 
made it an "activity" (Ivspyaa). See my edition of the Ethics, p. 3, 

2 The sense of the second question (15 b, 2-4) has been much dis- 
puted. I think that, if we read it with an emphasis on the first fuav 
and on ctvat, we shall see that it refers to the difficulty that arises when 
we predicate "being" of "one," that Is, when we speak, not merely of 
TO i> ?v, but of T> eV 6V. When we do that, the One at once seems to 
become two. That is a chief crux of the Parmenides. 


This section serves to link the Philebus to the Par- 
menides. At the beginning of the latter dialogue, the 
question of the One and the Many, so far as it refers to 
the predication of opposite attributes, and to the relation 
of whole and parts, is disposed of by the participation of 
sensible things in the forms, and it is then shown that the 
real difficulty lies in the union of One and Many in the 
forms themselves. If we say that the One is> it seems to 
become two on our hands ; while, if we say that sensible 
things participate in it, it is either broken up into parts 
and so becomes infinitely many, or the whole form must 
be present in each of the participants, so that we have an 
infinite number of ones alongside of the one One. No 
direct solution of this difficulty is given in the Parmenides, 
but a hint was thrown out that a solution was possible. 
We shall see that the Philebus puts us on the way to it. 

245. The difficulty that a thing turns into a one and 
many whenever we speak of it, really pervades all state- 
ments (Aoycu) or propositions we can make about anything 
whatsoever. It is " an affection of propositions in our 
minds (<n/ ^M^) that never dies nor ages." It is this that 
gives rise to all eristic disputation, and we cannot get rid 
of that till we have formed a sound theory of it. The 
only way to reach one is a way of which Sokrates has 
always been a lover (e/)acrT7fr), though it has often left him 
stranded, and it is the way in which all inventions and 
discoveries in the arts have been made. It is this. 

The gods once revealed to mankind, and the ancients, 
who were of a higher nature and nearer to the gods than 
we are, have handed it down as a tradition, that everything 
we say at a given moment (aei) is consists of one and 
many, and has Limit and Unlimitedness innate in it. What 
we have to do, then, is first to find a single form ($ea) in 
the thing we say y, and then to look in that for two 
subordinate forms, or three, or whatever number there 
may be. After that we must look at each of these new 
units and see how many forms are in them, until we are 
able to say of the original unit, not only that it is one and 


many, but also how many it is. We must not predicate 
the Unlimited (r^v rod aTrelpov ISeav) of the manifold, before 
we have gained a clear image of the number which is 
intermediate between the Unlimited and the One. Then, 
and not till then, may we give it up and let the manifold 
slip into the Unlimited. That is the genuine revelation 
of the gods, but the wise men of to-day are both too quick 
and too slow in setting up a One and a Many, and the 
middle terms (ra /xeVa) escape them. That is just the 
difference between dialectical and eristical discussion 
(15 d 17 a). 

Voice, for instance, is both one and many, but to know 
that does not make you a " grammarian " (phonetician). To 
become that, you must know also how many and of what 
nature the indefinite manifold is. In the same way, he is not 
a musician who can only say of a note that it is high or low 
or of the same pitch (as the keynote) ; he must know also 
how many intervals there are and of what nature, and what 
are the terms (Spot) of the intervals (i.e. the numbers, such 
as 12, 9, 8, 6, which express them), and how many scales 
these give rise to. Further, he must know to how many 
rhythms and metres the motions of the body when measured 
by numbers give rise (17 a 17 e). 

Just in the same way, when we have to start from the 
side of the Unlimited, we must not go straight to the One, 
but must carefully note the number of the intermediate 

If we start from sound, which is unlimited, we find first 
that there is a certain number of vowels, and then a certain 
number of liquids (^eo-a) and a certain number of mutes, and 
considering all these we bring them under the single unity of 
letters (crro^x^)' Then, and not till then, do we see clearly 
that the art of grammar has letters for its province, and not 
merely sound (18 a 18 d). 

A good example of the premature introduction of the 
Unlimited is afforded by the early Pythagorean treatment 
of the scale. If we were right in holding that they only 
determined the intervals of the fourth, the fifth, and the 
octave, referring all the internal divisions of the tetrachord 


to the Unlimited ( 30), that is just the sort of thing 
Plato means here. It is the more likely he had this 
in mind that we know Archytas and Plato busied them- 
selves with this very problem of the division of the 
tetrachord. We must also observe carefully that we do 
not eliminate the Unlimited altogether, but reach a point 
where we can no longer introduce number. That, too, 
can be illustrated from the musical scale, where we come 
ultimately to intervals which cannot be expressed as the 
ratio of one whole number to another. So far as we have 
yet gone, there is a point where division must cease. 

246. To illustrate what he means by the Unlimited, 
Sokrates takes the example of cc the hotter and colder," 
and this enables us to elucidate his meaning with the help 
of ^the distinction between heat and temperature, a distinc- 
tion historically connected with the Pythagorean doctrine, 
since, as we have seen, u temperature " is a translation of 

If we consider the sensation or quality of heat, we see 
at once that it varies in intensity. Water may be much 
hotter than our hand or only a little hotter, or nearly as 
hot, or not nearly so hot. In other words, heat " admits 
of plus and minus " (TO jmSX\ov KCU fjrrov). On the other 
hand, these degrees of intensity are quite indefinite. We 
cannot attach any clear meaning to the statement that one 
sensation of heat is equal to another, or that one sensation 
of heat is the double of another. These considerations 
explain what Plato meant by " the dyad of the Great-and- 
Small," which was his own name for what he calls the 
Unlimited in the Phikbus. It is the possibility of indefinite 
continuous variation in both directions from a fixed point. 
The Limit, on the other hand, does away with this inde- 
finite " more and less." Its simplest form is " the equal 
and the double" ( and ), and in general it is everything 
which " has the ratio of one number to another or one 
measure to another." This is the conception of quantity 
as distinct from that of quality, and its chief characteristic 
is that it enables us to speak with perfect clearness of equality 


and of addition, the simplest form of the latter being " the 
double" What enables us to do this is the introduction 
of a unit, in terms of which we may measure degrees of 
intensity. We cannot attach any clear meaning to the 
statement that it is twice as hot to-day as yesterday, but 
we do understand what is meant by saying that 60 is twice 
30. That implies further that a zero of temperature has 
been fixed, all temperatures above which are plus and all 
below it minus. The conception of negative quantity is 
thus clearly formulated for the first time in the history 
of science. 

247. Aristotle tells us further that the Great-and-Small 
was identified with "not being." 1 This doctrine is not 
attributed to Plato by name, but we fortunately possess a 
fragment of Hermodoros 2 which leaves no doubt upon 
the subject and also suggests the explanation. He says : 

Those things which are spoken of as having the relation of 
great to small all have the " more and less," so that they can 
go on to infinity in the direction of the " still greater" and 
the c still less." And in the same way, the broader and 
narrower, the heavier and lighter, and everything which is 
spoken of in that way can go on to infinity. But what is 
spoken of as equal and at rest and attuned has not the " more 
and less" as their opposites have. There is always something 
more unequal than what is unequal, something more in 
motion than what moves, something more out of tune than 
what is out of tune. [The text of the next sentence is 
corrupt]. ... So that what is of this nature is inconstant and 
formless and infinite, and may be called "not being" by 
negation of " being " (/carol air6<pa<nv rov 

If we have read the Sophist aright, the meaning of this 
is plain. It is not meant that the indefinite continuum 
of the more and less is nothing, but rather that it is not 
anything. We predicate of it the significant negative term 
(a7ro<pacrt9\ cc not being," not a blank negation which has 
no meaning 

1 Phys. 1 92 a, 

2 See Simpl in P/iys. p. 247, 30^. (Diels) 

OY2IA 331 

248. From all this it appears that we shall have to 
assume a third "kind" in addition to the Limit and the 
Unlimited, namely, the Mixture of both. We see this 
both in Medicine and in Music, where health and "har- 
mony" are produced by the due mixture of the two. We 
see the same thing in climate ; for a temperate climate is 
produced by such a mixture. The same explanation may 
be given of all goodness whether of body or soul, beauty 
of body and order of soul, and indeed all good things are 
due to such a mixture (256 sqq.*). 

The thought here is obviously Pythagorean ; it is just 
the tuned string once more. But there is a fundamental 
change in the point of view. The Pythagoreans had 
identified the Limit with good and the Unlimited with 
evil, but here we are distinctly told that, so far as human 
life is concerned, good things are all to be found in the 
Mixture. It is just for that reason that the " mixed life/' 
which includes both Thought and Pleasure, is found to be 
superior, not only to the life of Pleasure alone, but also to 
the life of Thought alone. 

249. Closely connected with this is the new sense in 
which Plato uses the term " being 7 ' (ovcria) in this passage. 
The Pythagorean doctrine simply identified the Form with 
being and the Unlimited with becoming, but Plato dis- 
tinctly states that the Mixture alone is truly " being." 
The process of mixing is indeed a " becoming" (-yeWi?), 
but it is a becoming which has being for its result (yevea-is 
i$ ova-lav) and the mixture itself is being, though a being 
which has become (ye-ye^e'i^ ova-la). Just in the same 
way we are told in the Timaeus (35 a) that being (owr/a) is 
a blend of the Same and the Other. These are only 
hints, and there are no others of the same kind in the 
dialogues, where they would be out of place, but they 
supplement what Aristotle tells us in the most interesting 
way. As the form-numbers are themselves a mixture, it 
follows that even sensible things may be real in spite of 
the fact that they are mixtures. In other words, the 
mature philosophy of Plato found reality, whether 


intelligible or sensible, in the combination of matter and 
form, and not in either separately. 

250. There has been considerable discussion as to the 
"kind" to which the "ideas" or forms belong in this 
scheme. The traditional view was that they were repre- 
sented by the Limit, and that is, of course, in accordance 
with the earlier Pythagorean version of the theory. It 
would be quite correct to refer the forms of the Phaedo 
and the Republic to this kind. Professor Jackson, on the 
contrary, maintains that the forms belong to the Mixed 
kind, and we have seen that the forms were certainly 
regarded by Plato as a mixture. On the other hand, it is 
surely plain that the Mixture of the Philebus is the world 
of sense, and the forms must, therefore, be referred to the 
Limit. The difficulty arises, I think, from the fact that 
Plato refrains from giving his full doctrine on the subject in 
this dialogue. From the point of view here taken, the 
forms belong to the Limit, but that does not alter the fact 
that they themselves are in turn a mixture. In the sensible 
world, their function is to limit, but in the intelligible 
world they themselves appear as a limited continuum, as a 
blending of matter and form, of the One and the Indeter- 
minate Dyad. 

251. Now this new view of reality clearly implies not 
only the categories of Being and Not-being, Same and 
Other, but also that of Motion, which was already asso- 
ciated with these in the Sophist ( 211), and this not only 
in the sensible but also in the intelligible world. We 
could only explain the generation of lines, planes, and 
solids by the help of this category ( 239), and if the 
sensible world is also a mixture, there must be a cause of 
the Mixture. That will be a fourth "kind" (27 b), 
and we must now go on to consider what Movement 
implies. Unless we can give an intelligible account of 
this, we have failed to explain the world we know* 



The Soul 

252. It was his theory of Soul that enabled Plato to 
account for Motion. Apart from that, we should have 
nothing but a string of what we may best represent to 
ourselves as algebraical formulae. The early Pythagoreans 
had grasped the conception of Soul as something more 
than the mere ghost of popular belief, but their later 
tenet that the soul is an " attunement " of the body made 
them lose hold of it again. Sokrates had insisted on the 
reality and eternity of the soul ; but Plato was the first to 
attempt a scientific justification of this belief. It is signifi- 
cant that the argument which seemed decisive to him does 
not occur in the Phaedo, though Sokrates is made to state 
it in the Phaedrus. In that dialogue we are told (245 c) 
that what moves another thing, and is in turn moved by 
something else, may cease to be moved and therefore 
cease to move anything else ; but what moves itself will 
never cease to move. It is the source and beginning of 
motion (a/>x^ Screws). Now such a beginning can never 
have come into being ; for everything that comes into 
being must have a beginning, while this is itself a begin- 
ning. Nor can it have any end ; for, if it perished, every- 
thing would come to a standstill. Such a beginning is the 
soul ; for it is the self-moved (TO <WTO eauro /avow/), and is 
therefore without beginning and without end. 

2,53, If this doctrine occurred only in the Phaedrus> it 


might be set down as mythical, though, despite the enthus- 
iasm of the passage, the language is curiously technical and 
scientific. It might also be said that it only proves the 
eternity of soul in general or of the world-soul, not that 
of the individual soul. In fact, however, the phraseology 
of the Phaedrus remained in use, and the question of the 
"first mover" continued to be a fundamental one. All 
doubt on the point is set at rest by the perfectly matter-of- 
fact treatment of the subject in the Laws, where we have 
an indication of Plato's mature thought on the subject. 

He begins (893 b) by distinguishing ten kinds of 
motion, of which the ninth and tenth alone concern us at 
present. The ninth is the motion that can move other 
things but cannot move itself, and the tenth is that which 
can move both itself and other things. It is really, Plato 
says, the first, since it is the beginning of motion (apx*J 
/awfereo>?) to the other nine. Now we do not find motion 
of this kind in earth, fire, or water, but only in what lives, 
that is, in what has a soul ; and if we ask for a definition 
of the soul, we can only say that it is <c the motion which 
of itself can move itself J) (ryv avrrjv CWT^I/ SwajmGvrjv KWGLV 
tclvrjortv'). The other motions all belong to body, and soul 
is therefore prior to body (896 b). 

But, if soul is prior to body, it follows at once that all 
the attributes of soul, such as characters, wishes, reason- 
ings, beliefs, forethought, and memories are prior to the 
attributes of body, such as length, breadth, depth, and 
strength ; and, if this is so, soul alone can be the cause of 
good and bad, fair and foul, righteousness and wickedness, 
and all other such opposites. There are such things as bad 
habits and bad reasonings, so there must be at least two 
souls, one that does good and the other that does the 
opposite (896 e). 

This passage is generally supposed to assert the exist- 
ence of an evil world-soul as well as of a good one, but it 
is important to observe that this does not follow from the 
words of Plato. He does not say there are two souls, a 
good and a bad one, opposed to one another, but that 


there are not less than two. It is as illegitimate to infer 
that there is only one evil soul, as it would be to infer that 
there is only one good soul, and it is rather implied 
that there is a plurality of souls, some good and some evil, 
We shall see presently that there is one pre-eminently 
good soul, namely God, but there is no suggestion of a 
pre-eminently evil soul, and that view is expressly rejected 
in the Statesman (270 a). The main point is rather that, 
since evil exists, there must be a plurality of souls for 
evil as well as good must be caused by a soul, whether by 
one soul or many. That is the important thing. We can 
no longer refer evil to body or matter ; the philosophy of 
movement requires us to attribute it to soul just as much 
as good. 


254. Now, if we look at the motions of the heavenly 
bodies, we see at once that they must be caused by a good 
soul or souls, and indeed by the best, since they are the 
most regular of all motions. That is due to their circular 
character, which must have been given them by a good soul, 
since, if left to themselves, things do not move in a circle 
but in a straight line. 1 These souls are what we call gods, 
if there are many, or God, if there is one only, or one 
which is the best of alL It is in this way that Plato 
reaches what he believes to be a scientific proof of the exist- 
ence of God, and it is only when he has done this that he 
can explain the world. There can be no sort of doubt that 
Plato regarded this as the central thing in his philosophy, 
and we shall understand that just in proportion as we 
realise this fact. At the same time, we must note at once 

1 This was rightly insisted upon by the Platonist Atticus (2nd cent. A,D.) 
as the fundamental distinction between the theories of Plato and Aristotle. 
Aristotle made the circular motion (/cwAo^opta) natural to the heavens, 
while Plato held that it must have a cause. We call this cause Gravity 
and we know much more than Plato did of the way in which it acts, but 
we know no more than he did of its nature. Plato knew there was a 
problem here ; Aristotle denied that there was any. 


that, though he believes this line of argument sufficient to 
demonstrate the existence of God, it tells us no more about 
him than that he is the self-moving source of good motions. 
Even so this is something quite different from anything 
the earlier philosophers had meant when they spoke of 
God. The lonians had called fire, air, water and the like 
gods, but that only meant there were no other gods but 
these. Anaximander and Xenophanes had called the worlds 
or the World gods or God, but that was at most a sort of 
pantheism, as it was also with Parmenides. Belief in God 
was doubtless -part of the Pythagorean religion, but it was 
hardly a part of Pythagorean science. Plato brought the 
idea of God into philosophy for the first time, and the 
form the doctrine took in his mind was that God was a 
living soul and that God was good. So much as that, but 
no more, he believed himself to have established by strictly 
scientific reasoning. 

We must not assume, therefore, that Plato meant by 
God exactly what a modern theist would mean by the 
word. Plato's God is certainly a " personal " god, as we 
should put it ; for he is Mind (vows) existing in a living 
soul, but it does not follow that he is the "supreme being". 
We have seen (171) that Plato continued to lecture on the 
Good to the last, and it is clear that his deepest thought 
was expressed in this lecture, so far as it was expressed at 
all. The way in which one of his followers after another, 
including Aristotle himself, endeavoured to publish an 
authentic report of it proves that it was regarded as 
fundamental. The question that arises, then, is whether 
we are to identify God with the Good or not ; and, if we 
are not, what relation we are to understand God to have 
to the Good. This question is not so simple as it appears ; 
indeed, it is highly ambiguous. If it is asked whether the 
Good is to be identified with the conception of God as 
held by modern theists, the answer is that it is certainly 
included in that conception, though it by no means 
exhausts it. If, on the other hand, it is asked whether the 
Good is to be identified with the God whose existence 


Plato believed himself to have proved by the argument 
just explained, the answer must certainly be that it is not. 
The Good is not a soul, but a " form." That is just how 
Plato avoids pantheism, which he regards as equivalent to 

255. This conception is not without its difficulties, as 
Plato was well aware. In the Timaeus he says (28 c) " To 
find the maker and father of this universe is a hard task ; 
and, when you have found him, it is impossible to speak 
of him before all people." That is a sentence of un- 
questioned authenticity, and fully explains the enigmatic 
manner in which Plato speaks of the same difficulty to 
Dionysios (who imagined he had solved it) in the Second 
Epistle (312 e). It also explains why he never wrote or 
published the Lecture on the Good, and why in the Laws, 
which was written for publication, he always speaks of God 
and never of the Good, though the Laws must be con- 
temporary with that very lecture. The problem continued 
to be discussed wherever there was living Greek thought. 
Some later writers regarded the Good as the supreme God, 
and made the Creator of the world subordinate to him, 
and there were many other attempted solutions. The 
difficulty is, in fact, the source of the controversies which 
were ultimately settled by authority at the Council of 
Nicaea, though this did not prevent it from continuing to 
trouble the minds of original thinkers. That does not 
concern us here. All we have to make clear is that Plato's 
God is not a form but a soul, and that he is the self-moved 
mover of the best motions. The Good is not a soul, but 
it is independent of God, and even above him, since it is 
the pattern by which he fashions the world. 

It is equally certain that God is not the only self-moved 
mover but simply the best of them. No doubt the sub- 
ordinate gods of the Timaeus belong to the mythology of 
that dialogue, and we can hardly doubt that Plato was a 
monotheist. The question, however, of monotheism or 
polytheism was not an important one to the Greeks, and 
Plato might have admitted other gods, so long as they 


were strictly subordinate. The main point is that human 
souls, though inferior, exist just as truly as the divine soul, 
and that in this way Plato thought it possible to reconcile 
the existence of evil with the absolute goodness of God. 
Here too we are faced by a difficulty which continues 
to trouble mankind. Are individual souls in any sense 
created by God, or is their existence entirely independent 
of him ? In the Timaeus there is a hint of a possible 
solution of this question. We learn there that individual 
souls are indestructible, not in their own nature, but 
because to destroy what he has made is inconsistent with 
the goodness of God. How far such a solution would 
really express the mind of Plato cannot be determined till 
we have come to a conclusion about the principles on 
which the Timaeus is to be interpreted. 

The World. 

256. The Timaeus^ which was certainly written long 
after the Republic, professes to describe a meeting which 
took place the day after Sokrates repeated the conversation 
narrated in the earlier dialogue, and consequently two days 
after that conversation itself. That makes a busy three 
days, especially as the Timaeus was to be followed at once 
by the Critias, which Plato has left unfinished, and by 
the Hermocrates, which was never written at all. We learn 
for the first time in the Timaeus that the audience to 
which Sokrates repeated the Republic consisted of Plato's 
great-grandfather, Kritias, 1 Timaios the Lokrian, Hermo- 
krates, and an unnamed fourth person who is prevented 
by illness from being present the next day. It is not very 
profitable to speculate who he may have been, but it is at 
least certain that he was a Pythagorean ; for Timaios is 

1 See Appendix. It is made perfectly clear that this Kritias is not the 
Kritias who was one of the Thirty, but his grandfather, though the two 
are hopelessly confused by modern writers. He is a very old man, who can 
hardly remember what he was told yesterday, but remembers the scenes 
of his boyhood clearly (26 b). At that time the poems of Solon were 
still recent (sib). It seems clear to me that most of the poetical frag- 
ments ascribed to the younger Kritias are really his grandfather's. 


represented as his understudy and agrees to replace him. 
If a name has to be given, I would suggest that of Philo- 
laos, and I should explain his absence by the consideration 
that the Timaeus^ though certainly based on his system, in 
several points goes beyond what we can reasonably attri- 
bute to him. If that is so, we can understand the origin 
of the famous scandal that Plato plagiarised the Timaeus 
from the " three books " of Philolaos which had come into 
his possession. 1 

However that may be and I only offer the suggestion 
for what it is worth the elaborate mise en scene must 
surely have some significance. If Plato took so much 
trouble to attach the Timaeus to the Republic^ he must 
have meant the later dialogue to supplement the earlier in 
some way, and this must be connected with the startling 
fact that Sokrates begins by giving a recapitulation of the 
Republic which includes Book V., but ignores Books VI. 
and VII. altogether. We are not allowed to attribute 
this to an oversight ; for Sokrates asks Timaios whether 
the summary is complete, and receives the answer that 
nothing is lacking (19 b). This can only mean that 
the Timaeus and its projected sequels were intended to 
replace in some way the later books of the Republic. The 
fact is that the central books of the Republic do not, except 
in the matter of solid geometry, go materially beyond 
what Sokrates might have learnt and probably did learn, 
from his Pythagorean associates, and Plato now wishes to 
make a further advance. For the same reason, Sokrates 
is no longer the chief speaker. The new views, however, 
are introduced with great reserve and somewhat obscurely 
expressed, so that there has been much dispute as to the 
meaning of some of the most important passages. Plato 
does not forget that the dialogue is supposed to take place 
in the fifth century. 

257. The Timaeus professes to give an account of the 
creation of the world, and the question at once arises 
whether this represents Plato's own doctrine or not. It 
iJE. Gr. Wi4- 


is quite certain that Xenokrates and other early Platonists 
held it did not. The world, they said, was represented as 
having a beginning in time only for purposes of exposi- 
tion (&<W/caX/a? yapiv), just as the construction of a 
diagram may be the best way to exhibit the properties 
of a figure. Aristotle thought it necessary to argue 
against this principle of interpretation, and we may say 
that, on the whole, the Platonists regard the Timaeus as 
mythical, while the Peripatetics take it literally. That, 
however, is impossible for anyone who has grasped the 
central doctrine of Platonism. We can infer the existence 
of the soul and of God from the fact of motion, but we 
cannot give any scientific account of the way in which 
they act. The world of experience is only, after all, an 
image, and it belongs to the region of becoming, and 
we can therefore do no more than tell "likely tales" 
(encores Xo'yoi) about it. Cosmology is not, and cannot 
be science, any more than Theology or Psychology. It 
is only a form of " play " (TTCM&O). Science, in the strict 
sense, must be mathematical. And yet Cosmology is not 
mere play either, for our account of the world will be 
related to the truth in the same way as the world is 
related to reality. It will be truth in the making, just as 
the sensible world is the intelligible world in the making. 
The appropriate vehicle for half-truths of this kind is 
myth, and here we must note once more that myth 
expresses something lower than science, and not some- 
thing higher. That is fundamental for the interpretation 
of Plato. The matter is put quite clearly in the Timaeus 
itself. We are dealing with what is always becoming and 
never is, not with what always is and never becomes (27 d). 
The former is an image (eoccoy) of the latter (29 b), and 
the work of ordering the sensible world after the pattern 
of the intelligible is assigned to God. No description of 
this process can have a scientific character, for we are 
dealing with what cannot be an object of knowledge, but 
only of belief (agb-c), and knowledge is higher, not 
lower, than belief. 


258, We are first told that God found a visible mass 
moving in a disorderly fashion, and resolved to bring it 
out of disorder into order. If we ask why he did so, the 
answer is " He was good, and the good has never at any 
time a feeling of jealousy towards anything, so he wished 
everything to become as like himself as possible" (296). 
This he brought about by creating a soul of the world, into 
which he introduced mathematical and harmonic relations 

We note here, in the first place, the phrase "as like 
himself as possible." This reservation is called for because 
Mind (i/ow) is confronted by Necessity (avajKt]\ and 
cannot, therefore completely effect its purpose (47 e). We 
must, then, consider the "errant cause" (TrAaj/oyxeVj; ama). 
In particular, we must explain how the elements came into 
being. For these cannot be ultimate. So far from being 
" letters " (o-ro^em, elementd]^ they are not even syllables. 

The conception of Necessity to which we are here 
introduced is not by any means an easy one. It is 
certainly not what we call physical necessity, for we are 
told that it can be " persuaded" by Mind. We are even 
told that it is a cause, and a cause "subservient to" the 
divine. It is a " concomitant cause " (crwamoi/) of the good- 
ness of the world, which could not be realised without it. 
This idea is as old as the Phaedo, where the concausa as dis- 
tinct from the causa is defined as " that without which the 
cause would never be a cause" (99 b). We learn further 
that this "concomitant" or "subservient "cause is corporeal, 
and that most people make the mistake of confusing it 
with the true cause, explaining everything, as they do, by 
warming and cooling, rarefaction and condensation, and 
so forth. The true cause is Mind and Mind alone, and 
the corporeal is a hindrance as well as a help. Mind 
could do nothing without something to work on, but 
that of itself stands in the way of it carrying lout its 
purposes completely. We learn also that these secondary 
causes " are moved by something else, and then of 
necessity move something else," as contrasted with the 


primary cause, which is self-moved. That is to be under- 
stood izi the light of the doctrine of soul discussed above 
( 256). It may help the reader to appreciate the account 
Plato makes Timaios give of Mind and Necessity if he 
will compare it with the theory of Leibniz that this 
is the best of all possible worlds. The difference is that 
Plato regards his explanation as a myth, while Leibniz 
considered his to be an adequate solution of the difficulty. 
259. This purely mythical character of the cosmogony 
becomes still more evident if we consider its details. In 
particular, motion is ascribed to the disordered mass before 
the world has received a soul, and that is in flat contradic- 
tion to Plato's doctrine that soul alone is self-moved. 
Plutarch, one of the few Platonists who took the Timaeus 
literally, can only get out of this difficulty by the help 
of the evil world-soul supposed to be assumed in the 
Laws ( 256). That, according to him, is eternal, and 
is to be identified with Necessity ; only the good world- 
soul was created. But, even supposing Plutarch to be 
right in finding an evil world-soul in the Laws, there 
is certainly nothing said about it in the Timaeus^ and it is 
impossible to suppose it would not have been mentioned 
if so much depended upon it. Besides that, we have seen 
that Necessity is " subservient " to Mind. A similar diffi- 
culty arises when we consider what is said about Time. 
In the Timaeus it is spoken of as a " moving image of 
eternity" (37 d), and we are told that it comes into 
being " along with the heavens" (38 b), that is to say, 
after the creation of the world-soul, which does not, there- 
fore, take place in time. That gives us the explanation of 
the necessarily mythical character of the whole story. We 
can only think of motion as in time, for time is just the 
measure of motion. On the other hand, knowledge is of 
the eternal and not of the temporal. It follows that, 
when we have to speak of motion, our language is perforce 
unscientific and pictorial. It can only convey an " image " 
of the truth, since time itself is only u a moving image of 
eternity." This does not mean, as we shall see, that time 


is subjective, but only that we fail to grasp its true nature. 
It is really the continuum implied in the conception of 
motion, but that cannot be known in abstraction from 
motion itself. 

260. But, besides being temporal, the " errant cause " 
is spatial. This is also hard to express in words ; for 
space is apprehended neither by thought nor by sense, but 
by " a sort of bastard reasoning " (Xajio-jmS TLVL v66(p). It 
is a sort of " receptacle" (VTTO^O^) or cc nurse" (rtOvvri) of 
all things (49 a). To understand this, we must go back 
to the elements, which we have already denied to be 
primary. We see that they pass into one another by rare- 
faction and condensation, and it is safest not to call any of 
them "this," but only "such" (49 d). The only thing 
which can be called " this " is that " in which " (/ <) they 
all appear to arise and pass away (49 e). 

This may be illustrated by an example. If we were to 
make all sorts of forms out of gold and keep constantly 
changing them, the only answer to the question " what 
is that?" would be "Gold." We should not speak of 
the transient forms it assumed as "things" (o>? oi/ra) at 
all. It is the same with "the recipient of all things" 
(TO TravSexei), the matrix (eKjmayeiov) on which the forms 
are "impressed" (evrvTrovvrai) . It has itself no form, 
but remains always the same, taking on with complete 
indifference the forms that " pass in and out of it " (ra 
acnoVra /ecu e*oVra), and these in turn are " imitations of 
what is ever" (rS>v OVTCW aeJ //u/ttjyuara) . They are, in fact, 
the elementary triangles and their products the regular 
solids, and we know from Aristotle, though we are not 
told so in the Timaeus,thnt they are imitations of numbers. 
We must, therefore, distinguish three things, the Form, 
which is the father, the Recipient, which is the mother, and 
the offspring of the two (the Mixture of the Philebus\ 
which is the Corporeal. The Recipient is altogether form- 
less ; all we can say of it is that it is an invisible, all- 
receptive something, partaking in a mysterious way in the 
intelligible. It is, in fact, space (x< 


261, That the so-called "primary matter" of the 
Timaeus is space of three dimensions and nothing else is 
really quite certain both from Plato's own language on the 
subject and from the statements of Aristotle. Nor is there 
any occasion in the system for any other kind of cc matter." 
The " elements" of the corporeal are completely accounted 
for by the regular solids, and they in turn can be con- 
structed from the elementary triangles. Plato undoubtedly 
means to say that the corporeal can be completely reduced 
to extension geometrically limited. Indeed, he goes a 
great deal further than that, though he only gives us a few 
hints of his real meaning here. We do not perceive space 
at all by the senses ; we only infer it by a species of reason- 
ing, and that reasoning is a "bastard" one. It is "in a 
dream" that we say everything must be in a place and 
occupy a space (52 b), and when the elementary triangles 
are discussed, it is said that the principles (jox a which 
are higher than these God knows, and of men he who is 
dear to God (53 d). Space is only one aspect of Con- 
tinuity, and not an essential one. These considerations, 
however, take us beyond the mythology of the Timaeus, 
for which space is ultimate. 

262. The corporeal world, then, is in space and time, 
and for that reason it can only be described in mythological 
language. That does not, however, exhaust Plato's teach- 
ing on the subject. What we say of the world is not, 
indeed, the truth, but it may be more or less like the 
truth, and it is our business to make it as like the truth as 
possible. The boundary-line between the intelligible and 
the merely sensible is not a fixed one, and the sensible 
may be made progressively intelligible. It will, I think, 
be admitted that this is the doctrine to which all the 
dialogues from the Theaetetus onwards naturally lead 
up, and I believe we shall find proof that Plato held it. 
Unfortunately, however, his followers were not able to 
rise to this point of view, and Plato has been generally 
credited with an absolute dualism. Xenokrates confined 
the province of science to the things "outside the heavens," 


and made the heavens themselves the objects of belief 
(<Wa). They were intelligible by the help of astronomy, 
but they belonged to the sensible world as being visible. 
If this report does justice to him, he made absolute a 
distinction which for Plato was merely relative. At the 
same time, it is just possible that this report may be only a 
distortion of what we shall find to be the true Platonic 
doctrine. There is no doubt about Aristotle, however. 
It is certain that he introduced for the first time the 
fatal notion that the nature of the heavens was quite 
different from that of the sublunary world. It is this 
doctrine, generally known as that of cc the incorruptibility 
of the heavens," that the Platonist Galileo was chiefly con- 
cerned to disprove by calling attention to such phenomena 
as the new star in Sagittarius, and it is strange that Aristotle, 
who condemned Plato's perfectly legitimate separation of 
forms from sensible things, should himself be responsible 
for a much more questionable "separation" (x a) P L<T f Jt> ^ 
like this. There is no trace of anything like it in Plato. 
He certainly assigned an exceptional position to Astronomy 
and its sister-science Music in his philosophy, but that 
was simply because, in his own day, these were the sciences 
in which the intelligible was most obviously advancing at 
the expense of the merely sensible. Even in the Republic 
(530 d) it is hinted that there are more sciences of motion 
in space than these two, and we can see from the Par- 
menides (130 e) that a complete science would have to 
account for "hair, mud and dirt" as well as for the 
planetary motions. It is, however, from his astronomy 
alone that we can gain a clear idea of the relation Plato 
held to exist between the sensible and the intelligible. It 
would be out of place to discuss it fully here ; it will be 
enough to look at a single branch of it, and I shall select 
one which is commonly misunderstood. 1 

263. The great problem of the day was that of the 

1 This applies even to the recent discussion of it in Sir T. L. Heath's 
Aristarchus ofSamos, which in other respects is an excellent guide in such 


planetary motions. For the senses these are hopelessly 
irregular, and that is probably why we hear in the Timaeus 
of the cc errant cause" (7rXai/co//eV>; am'a). In the first 
place, since the paths of the planets are oblique to the 
equator, their apparent courses are spirals (We?), not 
circles. In the next place, Mercury and Venus at one 
time travel faster than the Sun, so that they get in front 
of it and appear as morning stars ; at another time they 
lag behind it and appear as evening stars. In fact, these 
three bodies are always " overtaking and being overtaken 
by one another" (38 d). The other planets behave even 
more strangely. Sometimes they seem to accelerate their 
velocity so as to appear stationary among the fixed stars 
or even to get some way ahead of them ; at other times, 
they are retarded and seem to have a retrograde motion. 
There is a further irregularity in the Sun's annual course. 
The solstices and equinoxes do not divide it into four equal 
segments as we should expect them to do. 

Now this irregularity cannot be ultimate. If we ask 
why not, the only answer is that the Artificer created the 
world on the pattern of the Good, and disorder of any 
kind is opposed to the Good. That is the ultimate ground 
of the rule that hypotheses are not to be needlessly multi- 
plied. The postulate of simplicity and regularity which 
still guides scientific research is at bottom teleological, 1 and 
we probably come nearest to Plato's thought about the 
Good if we say that, according to him, reality must be a 
system. There is something to be said, however, for 
his simpler way of expressing this. At any rate, it does 
not admit of doubt that Plato conceived the function 
of Astronomy to be the discovery of the simplest hypo- 
theses which would account for the apparent complexity of 
celestial phenomena. We know as a fact that he pro- 
pounded the solar anomaly as a problem to his scholars 


1 It is worth while to note that this term is derived from 
" complete," not immediately from T-Xos. It has no implication of an 
external end. 


264. Now we know further that Eudoxos invented a 
beautiful hypothesis, that of concentric spheres, to account 
for all these irregularities on the assumption of the earth's 
central position, 1 and we know also that Plato did not accept 
his solution as satisfactory. The assumption of twenty- 
seven spheres did not seem simple enough, and fuller study 
showed that still more were required. Kallippos added to 
their number, and Aristotle had to add still more. Finally, 
the concentric spheres were replaced by eccentric spheres 
and epicycles, and what we call the Ptolemaic system was 
the result. Besides this, Aristotle transformed the geo- 
metrical hypothesis of Eudoxos into a mechanical system 
of material spheres in contact with one another, and all 
that arrested the growth of a true astronomy for nearly 
two thousand years. 

265. Plato, on the other hand, saw clearly that the 
geocentric hypothesis was the source of the trouble. The 
later Pythagoreans had taught that the earth revolves round 
the Central Fire, and it was in this direction that a solution 
was to be looked for. Here again we have direct first- 
hand evidence. Theophrastos (who came to Athens before 
the death of Plato, and was almost certainly a member of 
the Academy) said that " Plato in his old age repented of 
having given the earth the central place in the universe, to 
which it had no right." 2 This is unimpeachable testimony, 
and no interpretation which ignores it can be accepted. 3 It 
does not follow from it, however, that Plato adopted the 
heliocentric hypothesis. 

1 For a clear account of this, see Heath, Aristarchus of Samos, pp. 190 


2 Pint. Quaest. Plat. 1006 c: Geo^pao-Tos Se /cat Trpocrto-ropet r<j> 
IlXarcovt TrpzcrpvTepQ yei/oju-evy /xera/xeA-eii/ a>s ov TrpocrrjKOiwav ttTroSotm 
T l? 7V r *) v JK&TIJV x^P av T v 7rai/T s- I n the Life ofNuma, u, Plutarch 
says, doubtless on the same authority : TlXdrtova. <acri Trpear/SvTrjv 
yevo/xei'OV Siavei/o^cr^cu ircpl r/js yf)<s <os Iv Irep^t X^PV / <a-^ o r Twcr^s, 
r^v Se /A(n?v KCU Ki^twrcxr^v cre/xj> rtvt Kpeirrovi TrpocryKOvcrav. 

8 Sir T. L. Heath (p. 186) says Theophrastos got the statement "from 
hearsay." No doubt, but he probably heard it from Plato himself, and 
certainly from his immediate disciples. 


266. Now there is a sentence in the Timaeus (40 b) 
which can only refer to the same doctrine, if we adopt the 
best attested reading. 1 The only admissible translation of 
this is cc earth, our nurse, going to and fro on its path round 
the axis which stretches right through the universe." The 
choice of a word which properly means " to go backwards 
and forwards" 2 is specially significant ; for it is just that 
aspect of the terrestrial motion which accounts for the 
apparently retrograde motion of the planets. This is enough 
for our present purpose, and I do not propose to discuss 
here the vexed question of whether the heliocentric 
hypothesis was mooted in the Academy or not. I believe 
it was, but in any case Aristarchos of Samos, who did pro- 
pound it, must have got his inspiration from the Academy 
and not from Eudoxos. 

267. Now let us see what light all this throws on 
Plato's philosophical position. In the first place, it is the 
phenomena of the visible heavens that furnish the problem 
for solution, and the assumption throughout is that it is 
possible to give an intelligible account of these. There is 
no attempt to shirk the difficulty by referring the irre- 

1 This is : yfjv Se rpoffiv p,V rjfAtrepav, iAAo/>ievr;v Se rrjv irtpl rov 
Bta Travrbs TroXov Tera/uVov. Everything here depends upon the word 
rrjv, which is quite distinctly written in Par. A, though omitted in all 
printed texts before my own. It can only be explained on the principle 
of rip (sc. 6Sov) ? and we must " understand " TrcptoSov or Trepifopdv. No 
"scribe" could have invented such a reading, which is also that of at 
least one other first-class MS. It is true that Par. A has tX.\opvr}v for 
iX.X.ojj.ev'iqv, but that is an everyday confusion, and the agreement of the 
MSS. of Aristotle, Plutarch and Proclus with other Plato MSS. turns the 
scale of evidence. 

2 The verb XAe<r0cu (which cannot be etymologically connected with 
etAAecr0at) has no other meaning than this in classical Greek literature. 
It is used by Sophokles (AnL 340) of ploughs going backwards and 
forwards in the furrow, and Xenophon (Cyn. 6) speaks of Kvvts 
e^iXXovorat ra i\^ going to and fro till they find the scent. If 
Apollonios Rhodios confused iAAo> and crAAa>, that proves nothing* 
Aristotle certainly understood the word to mean motion of some sort 
(de Coelo, 296 a, 5), and this is confirmed by the use of the present 
paniciple. It is quite incredible to rne that Aristotle should have mis- 
understood or misrepresented Plato's teaching on a subject like this. 


gularity of the planetary motions to the shortcomings of 
the sensible world, or to " matter " or to an evil world-soul, 
as popular Platonism did later. Nor is there any attempt 
to represent the phenomena as illusory ; on the contrary, 
the whole object of the inquiry is to "save" them. The 
appearances remain exactly what they were, only we now 
know what they mean. The gulf between the intelligible 
and the sensible has so far been bridged ; the visible 
motions of the heavenly bodies have been referred to an 
intelligible system, or, in other words, they have been seen 
in the light of the Good. If we ask why they should 
appear to us as they do, the answer must be on the same 
lines. It is because we are placed on a spherical earth 
which revolves round the axis of the universe, and that is 
because it is good that we should be so placed, though we 
cannot clearly see why in the present state of our know- 
ledge. That, I take it, is how Plato laid the ghost of the 
two-world theory which had haunted Greek philosophy 
since the time of Parmenides, and that is what he meant 
by saying that the sensible world was " the image of the 
intelligible." He had shown already in the Sophist that 
to be an image was not to be nothing. An appearance is 
an appearance, and is only unreal if we take it for what it 
is not. 


268. The account just given of Plato's mature philo- 
sophy is of necessity meagre and in a measure hypothetical. 
As to that, I can only say that in this case the phenomena 
to be t saved " are the writings of Plato himself and the 
statements of Aristotle and others who knew him, and the 
only proof or disproof the hypothesis admits of is its effi- 
cacy in accounting for them. It cannot be otherwise tested. 
Personally I have found this hypothesis efficacious during 
a course of Platonic study extending over twenty years at 
least, I claim no more for it than that, and also no less. 1 

1 It is nearly a quarter of a century ago that I found the current views 
of Sokrates and Plato leading me into a hopeless scepticism and resolved 


I do not pretend to impose my conclusions on the reader, 
who must make the experiment for himself. He will 
certainly find it worth while. 

There is another point still. It must be admitted that 
Plato's immediate followers fell very far short of the ideal I 
have attributed to their master. Aristotle was impatient 
with the mathematical side of the doctrine and did not even 
trouble to understand it. The result was that this did 
not come to its rights for nearly two thousand years. Even 
those men who were really carrying out the work Plato 
began felt bound to put their results in a form which 
Aristotle's criticism would not touch. The Elements of 
Euclid are a monument of that position. 1 Xenokrates 
confused Plato's philosophy of numbers with his philo- 
sophy of motion, and defined the soul as a " self-moving 
number." Speusippos held that the Good was not 
primary, but only arose in the course of evolution. The 
Neoplatonists did more justice to Plato's doctrine of the 
Good and of the Soul, but they failed to remember his 
warning that the detailed application of these could only 
be <c probable tales " in the actual state of our knowledge. 
Yet these very failures to grasp Plato's central thought 
bear witness to different sides of it and justify the attempt 
to reconstruct in such a way as to explain how it could be 
misunderstood in so many different ways. After all, these 
"broken lights" are also among the phenomena which 
have to be cc saved," and for this reason many sides of 
Plato's philosophy will only appear in their true light when 
we have seen how it fared in the hands of his successors, 
and especially in those of Aristotle. 

to see what could be done with the hypothesis that Plato really meant 
what he said. Since then I have edited the whole text of Plato, and an 
editor necessarily reads his text through minutely many scores of times. 

1 Perhaps the most significant touch is that he calls the axioms KOIVCU 
WVQUU or " innate ideas." That is a Stoic formula which enables him 
to avoid discussing the true nature of hypothesis. 



5 rr w n 
^G fi fi VM 

S n 



i-l y 



gj Q 



PH V i-I 

2 ci c ^ 



\ "* 

\ i i 

co v ~ x ^ O 



\ * 

p^ "^ 



. g o O 



l~ ^4*13 



\ & 

". c ^ 



\ * 

^ O cti P- 

j/T u flJ T3 

T> c r2 qj 



2 23 

a^l 15 

c o ^ 



^ H ft 


S . O !r5 

4) pW O 



s g 


^ o 


p i p. 

"C^O a, 






o c rt 
n^ w oj 










^ S. (^ 











"C D C^ 





W S 5P 

*tj .S <D 





g^ s^ 















j-[ ^ ^ n3 


WT3 C P 



"t S ^~ 













J^-2 u 


a g a" b ri 




'a ^11 

w a) l ~ l T 


5 H o* 2 " ^ 
no 3 ^ ft M CX 
r! -< P5 o Q r/ 
W J >^j CO g j3 g 
Q ^ q ^ 




^ d a a_ 

C 5j ^^ 3 

This table of P 
conclusions stated i 





S M fa 

&J 3| 


NOTE. Alexa 
former was celebra 
It will be observed 
for regarding the si 


Abaris, 40. 
Abdera, 194*. 
Academy, 213^., 303. 
Achilles and the tortoise, 84. 
Adeimantos, son of Ariston, 206 sq. 
Aether (aW-fjp) 21, ( = Fire in Anaxa- 

goras) 78. 

Aggregation, states of, 27. 
Aigospotaraos meteor, 80. 
"Air" (U. mist), 21, 25, 39, 44, 51, 

67, 72, 95- 

Air (atmospheric), 72, 78. 
Aischines against Timarchos (i) I39 3 , 

(1. 175) 187- 
Akoumenos, 190. 
Alexander the Great, 295. 
Alkibiades, 137, 138^., 141, 150, 187, 


Alkidamas, 3oo 2 . 
Alkmaion, 50, 75, 
Analysis, 219^. 
Anaxagoras, 76-81. 
Anaximander, 22-24, 
Anaximenes, 24-25. 
Andokides, 189^. 
Anthropomorphism, 29, 35. 
Antichthon, 92. 
Antisthenes, 172, 251 jy., 282, 
Anytos, no, 180, 186, 187, 188. 
Apocalypses, 65. 
Apolaustic life, 42. 
Apollo, 40. 

Appearances, saving, n. 
Arabic figures, 54, 
Archelaos of Athens, 124^., 147, 

Archimedes, 199, 322*. 

Archytas, 297, 323. 

Aristarchos of Samos, 348. 

Aristeas of Prokonnesos, 40. 

Aristeides the elder, 128. 

Aristeides the younger, 138. 

Aiistokles, 23O 1 . 

Aristophanes, 123, 142 s$., 144 sq. 9 
184 sq. 

Aristotle, II, 72, 345, 350; on Atom- 
ism, 95; on Sokrates, 157*; on 
Plato, 178, 312 sg?.; Categories (ga, 
8) 326 2 ; Prior Analytics (67*2, 

21) 158; Sophistici Elenchi (1650, 

22) 108, (i;8, 36) 260; Physics 
(I92, 6 j^.) 330, (203, 13) 53 2 > 
(250^2, 20) 114 sq. t (265<5, 25) 27 1 ; 
DC cado (2964, 5) 34S 2 ; De genera- 
tions (335^, 10) 1 66; Metaphysics 

(9870, 20) 315, (9870, 22) 157, 
(987^, I) 157, (987^, 29) 315, (989*, 

5jy/.) 26 1 , (9920, i) 322 *- 3 , (992^, 
32) 3I2 1 , (998, 2) 114, (1014^, 16) 
27 1 , (1043^, 5^///,) 251^;., 
9W-) 3i3 x > ( I0 7^ n) 157, 
17) 317, (1078^, 21) 157, (10784, 
30) 165, (io8o, 12^/0314, (loSia, 
35) 314, (10833, 33) 314, (io863, 3) 
317, (io9ia, 4) S 2 ^ (1095^, 8) 
53 a ; 7^7(145^,25) 183. 

Aristoxenos, 41, 48, 49, 87, 124, I29 fl , 
I34, 153, 221, 300 3 . 

Arithmetic, 84, 224, 323. 

Arteries, 77 1 . 

Astrology, 6, 7 a . 



Astronomy (Babylonian), 7 sqq. ; (Aca- 

demic), 225 sqq y 
Athens as the meeting-place of Italic 

and Ionic philosophy, 85, 119, 123, 

I3 2 . 214. 
Atomism (#, Leukippos and Demo- 

kritos), 86 ; (origin of), 95 ; and Pytha- 

goreanism, 97. 
Atoms, motion of, 96. 
Atticus the Platonist, 335 1 . 
Axiochos, 190. 

Babylonian astronomy, 7 sqq.> 19. 
Being (ova-ta) and becoming (ytve<rt.s) t 
90, 155 sqq., 159, 162, 224, 284, 

Biology, 24. 
Blend (/cpacris), 48. 
Blood, 75. 
Body, 31. 
Boundless, 22, 
Bryson, 254, 263. 

Campbell, Lewis, 2I2 1 , 252. 
Carthaginians, 294^., 300 .537. 
Categories, 247, 257, 283^^. 
Catharism, 41. 
Charmides, 138, 207, 210. 
Conceptualism, I54 1 , 258, 317, 
Condensation, v. Rarefaction, 
Constitutions, 292^. 
Continuity, 83, 84, 90, 114^., 19$ sg., 

320 sqq. 
Copernicus, 5. 
Cosmogony, 4, 18. 
Crete, 17, 30, 40. 
Croesus, 19. 
Cujas, 304 1 . 
Cyprus, 295 2 . 

Damon, 190. 
Darkness, 51^., 60. 
Delios of Ephesos, 295 1 . 
Delos, 30. 
Demeter, 30. 
Demiourgos, 49. 

Democracy, 293. 

Demokritos, 193-201 ; and Protagoras, 

116, I94 2 , 197 sg.y 244 1 . 
Diagoras, 76. 
Dialectic, 134^., 164, 228^., 263, 


Dialexeis, 23 1 2 . 
Diapason, 47. 
Dikaiarchos, 87, 153. 
Diogenes of Apollonia, 123, 145. 
Dion, 211, Z^sqq. 
Dionysios I., 211, 294^. 
Dionysios II. , 294 sqq. 
Dionysos, 30, 31. 
Division, 220. 
Dodecahedron, 6, 55, 83. 
Dramatic and narrated dialogue, 234. 
Dropides, 207^. 
Dualism, 89. 
Dyad, indeterminate, 314, 320^. 

Earth, Mother, 26. 

Earth (shape), 20, 23, 25, 36, 44, 72, 
So, 100 ; (place), 92, 347 ; (inclina- 
tion), 100 ; (as an " element"), 26. 

Echekrates, 152. 

Eclipses, 8, 18, 19, 24, 44, 60, 80, 92 

Ecstasy, 31. 

Education, 305-311. 

Effluences, 75, 119. 

Egyptian mathematics, 5 sqq.^ 210 $q. 

Elea, 33, 64. 

Eleatic Stranger, 237, 273 sq* 

Elements (v. <rroexa), 26, 6l, 69, 72, 
77, 78, 87, 88, 99. 

Empedokles, 43, 71-75, 119. 

Enlightenment, 32 sqq. 

Epameinondas, 219, 300. 

Epicharmos, 63. 

Epicurus and Epicureans, 23, 36, 96, 

Eristics, 172, 231, 242, 254, 263, 273. 

Eros, 138 sq. 

Eryximachos, 190. 

Euclid (axioms) 350, (I. 47) 40, 54, 
(II. 11)55, (XIII.) 219. 



Eudemos of Cyprus, 298. 
Eudemos of Rhodes, 20. 
Eudoxos, 24, 199, 214, 263, 324, 3252, 

Eukleides of Megara, 152, 161, 210, 

230 j^., 235. 
Eurytos, 53*, 90 sq. 
Evil, 334 sq. 
Evolution, 24, 75. 
Experiment, 10, 7 2 > 7^ 
Explanation, 10. 

Figurate numbers, 54. 

Figures (v. etfy, IMu), 49> 50. 5*, 52, 

54, 88 sqq. 

Fire, 60, 6 1 ; central, 92. 
Flux, 61. 
Fluxions, 322. 
Force, 70. 

Form and matter, 44, 56. 
Forms (v. cldy, IStai), 154 sqq,, 255 sqq. 
Fractions, 85. 

Galileo, 24, 220, 345. 

Geocentric hypothesis, 92. 

Geometry, plane, 20, 225. 

Geometry, solid, 213, 225. 

Glaukon, son of Ariston, 206 sq. 

Gnomon, 7, 53. 

God (gods), 23, 25, 28, 29, 32, 35, 63, 

Si, 117, 123, 169% 289, 335 W- 
Golden section, 55- 

Good, the, 169 sq., 221, 231 sq., 336 sq. 
Goodness, 109 sq, 9 170 sq., 173 sq., 

175 *? 

Gorgias, 119 sqq, 
Great-and -small, the, 314, 320 sqq., 

Gymnastics, 306 sq. 

Harmonic mean, 48. 

Harmonics, 227 sq. 

** Harmony" (&pfjt,ovta), 45; of the 

spheres, 56. 
Health, 50. 
Heart, 73, 75- 

Hekataios, 22. 

Hellenes and barbarians, 34, 218. 

Herakleides, 297. 

Herakleitos, 57-63. 

Herakles, 1 1 8, 121, 

Hermodoros the Platonist, 205, 210, 


Hermodoros of Ephesos, 58. 
Hermokopidai, 189 sq. 
Heiodotos, 106 sq. t (i. 74) 18, (ii. 109) 

7, (iii. 38) 107, (iv. 95) 107, 

(v. 28) 17. 
Hesiod, 28, 34 sq. 
Hiero, 33 
Hipparchos, 7. 
Hipparinos, 300. 
Hippasos, 55, 87 sq. 
Hippias, 1 1 8. 

Hippokrates of Chios, n8. 
Hippokrates of Kos, 10, 32, 33, 86, 

lor, 163. 

Hippon of Samos, 123 sq. 
Hipponikos, in 1 . 
Homer, 28, 34 sq. 
Homeric Hymns, 30 sq. 
Homo mensura, 114 sq. 
Hyperboreans, 30. 
Hypotenuse, 40. 
Hypothesis, 149, 162, 222, 224, 229, 


Images, 89, 276 sqq., 286 sqq., 315. 

Incommensurability, 54> 9Q> **4 3*o 

Indian science, 9. 

Indivisible lines, 322. 

Infinitesimals, 199, 320, 322. 

Infinity, 86, 

Injustice, 22, 48, 106. 

Intelligible and sensible, 316, 3401^. 

Intervals, musical, 45 sqq., 328 sq. 

Ionia, 214, 

Irony, 132. 

Isokrates, 107, 172, 215 sqq., 273, 295 2 , 
(I0,i) 171, (10,2) 113, (10,3) 120, 
(10.5) 172, (XI. 5) 138', i SoP, 
(13.1,3) 172. 



Justice (cosmological), 22, 61, 106. 

Kallias, no. 

Kallikles, 120^7., 186. 

Kallippos, 299. 

Kebes, 151 sq. 

Kepler, 38. 

Klepsydia, 72. 

Korybantes, 41. 

Kratylos, 241 sq., 315. 

Kritias, son of Diopides, 208, 338 1 . 

Kritias, son of Kallaischros, 138, 187, 

209 sq. 

Kroton, 39, 41. 
Kylon, 39. 

Law and nature, 105 sqq., 117, 122. 

Lawgivers, 1 06. 

Leukippos, 92, 94-101. 

Likeness, v. Images. 

Limit and unlimited, 44, 51, 327 sq. 

Lives, the Tbce, 42. 

Love and stufe, 72 sq. 

Lydia, 17, 19. 

Lyre, 45. 

Lysimachos, 128. 

Lysis, 219, 300, 

Maieutic, 139 sq. 
Man is the measure, 114 sq. 
Materialism, 279 sy. 
Mathematicians and Akousmatics, 88. 
Mathematics, 38; . Aiithmctic and 


"Matter and form, 27, 44, 56, 68. 
Mean, 48, 56. 
Measure, 114 sg* 

Medicine, 41, 49 ty'/., 88, 123 sqq. 
Megarics, I34> *5 2 > 230 sq., 235, 242, 

254, 272, 273^7., 277 syq. 
Meletos, 180. 
Melissos, 85 sq. , 95. 
Menon's latrika, 88, 123. 
Metapontion, 39, 40, 
Metempsychosis, 43. 
Might is right, 121. 

Miletos, 17, 28. 
tfind, 79, 123. 

Mixture (v. Blend), 31, 74, 76. 
Moon, 24, 36, 60, So, 92, 226. 
VTore and less, the, 329 tqg. 
Motion, 68, 69, 79, 84, 98, 245, 333, 


Music, 41, 45 sqq., 306. 
IMyrto, I29 2 . 
Mysteries, 12, 189 sqq. 
Mysticism, 168. 
Myth, mythology, 3 sqq., 167 sqq., 183. 

N"ariated and dramatic dialogue, 234 sq. 

Nature and law, 105 sq. 

Necessity, 341 sqq. 

Negative quantity, 330. 

Not being, 274 sqq. , 330, 

Numbers, 51-54, 83, 85, 89, 312-324. 

Oinopides, 80. 

One and many, 264 sqq. , 326 sqq. 
Opposites, 22, 48, 49 sq., 79, 88. 
Oibits, planetary, 24. 
Oiphicism, 31, 32, 59, 71, 130 sq., 

Paimenides, 51, 63-68, 133, 236. 
Participation, 165, 255 sqq., 282 sqq., 


Pentagon, regular, 6, 55. 
Pentagram, 55. 
Pentalpha, 55. 
Perikles and Anaxagoras, 76, 81 ; and 

Zeno, 82 ; and Melissos, 86. 
Peisia, 295. 
Phaidros, 190. 
Pherekydes, 4, 18, 26, 40. 
Philip of Opous, 301. 
Philistos, 295, 296. 
Philolaos, 87, 92, I53> 339- 
Philosopher-king, 218, 291 sg. 
Philosophy, 3 sqq., 4 2 > 2I 5' 
Phleious,- 152. 
Pindar, 107, 121. 
Planets, 7, 8, 226, 346 sq. 



Plato, 182, 205-350; Euthypkro (20) 
1 80, (&sg.) 184, (5^)i54\ (&0 183, 
(6r, e) I54 1 ; Apology 180 *#., (i70 
128, (260 i8o 2 , (sirf) i84 2 , (33) ^S, 
187, (34/0 206, (38^) 210, (39rf) 187; 
(45*) ISL (52*) I2 4 l , ($26) 128 ; 
, (58^)152, (59*) J 5i> 210, 
153 (65*, 66 ^> ) 42 1 , (730 IS 8 , 
(740 3 iS 1 , (763) 1 55, (7&0 158, (82) 
174, (S 4 160, (84*) 160, (850 160, 
(850 160, (863) 50, 93, (883) 161, 
(880 i6i,(88</),93, X 53 (89^-) l6l > 
(960 jf .) 132, (96*) 75* I2 5, (960 *35> 
(973) 125, 162, (98^) 80, (99*0 *6a, 
(99 W-) 315. (ioo3) 155, (iooO 164, 
(loirf) 164, (xoiff) 162, (I02) i53> 
(1023) 255, (io8) 201 ; Cratylus 
(3893) I54 2 , (400^:) 131 ; Theactetus 
237-253, (142^) 152, (i43) 236, 
235, (I48fl) 323, (15^) X 46, 
113, (iSoff) 86, (2iOfl?) 238; 
218, 273-289, (223^) 108, (242^) 64, 
(2480) 91 ; Politicks 290-294, (262</) 
218, (270^) 335? Parmtnides 253- 
272, (128^ 82, (I28a0 163, (1300) 

134, (130*) 155* (ISQ^O 9i (I3S<0 
134; /%z7tf3^, 324-332 ; Symposium 
i823) 139, (202tf) 251, (2150) 141 $?-, 
(220^)130, 137, (2150)141; Phaednts 
(227^ ^.) 185, (231^ I39 3 , (245^) 
333, (247^ i67 ( 2 5) J 40, (267^) 
r34, (2680) 190, (2790) 216; Char- 
mides 176, (1330) ^3^, ^^ (*5&*) 
207 ; Zrt^J 176 (178/1 sqq.) 146, 
(i8l3) 137, Euthydemus 135, (290^) 
229; Protagoras (3090) in, (3*10) 
in 1 , (3Xfl*)228 l , (315^ m 1 , (317/0 
109, (JI70 m, (36i<?) 134; Gorgias 
(456^) 175. (484*) I2I (5Q4 
177, (52**) 186; Meno (72^) 
(760 119, (790 HO> a S x > 
130, (863) 157, (9^ W*) IIO > 
in, 112, (940 186, (97*) 173. (98) 
174; ffippiasmaior(*&2e) in ; ^z)> 
pias minor, 175; ^public (b in 1 
(3320 176, 

(435*0 224, (47&*} 165, (504/0 224, 
(5050*^.) 169, (5100 229, (51 13) 162, 

22S 2 , (5200 245, (5243-5253) 224, (527a 

sqq.) 225, (528^^.) 225, ($2(}dsqq.) 

226, (5333) 229, (534)I59 1 ; Timaeus 

338-349, (19 a-d) 237, (280 337, (350) 

331, (403) 348, (483) 88, (510 155, 

(520 99 (5&*) 51 *; CnVwj (norf 

sqq.) 223; Z0WJ 301-311, (636^ sq.) 

138 J^., (6560 210, (7090 219, 302, 

(722^ sgq.) 301, (7340 294, (747*) 

211, (7473) 6, (7560 294, (8033) 303, 

(8193) 211, (8i9flO 310,. (8600*) 171, 

(8890 122, (SgiO 27 l , (8933-8960 

334; Epinomis (9873) 8, 

225, 322; Epistles 205 jy.; ii. 

337, (3140 212; iii. (3160) 301; v. 

(3220) 209; vii, viii., 300; vii. 

(320^, 32i3) 299 1 , (3240) 211, (3243) 

209, (3253) 210, (3260) 218, (3285) 

294, (341 c-d) i 1 , 221, (3450 297, 

(3530 3oo; xiii. (3^0 263, (361*) 


Pluralism, 69 sqq. 
Point, 83, 84. 

Polykrates of Samos, 34, 38. 
Polykrates the sophist, 1 38 1 , 150, 172, 


Polyxcnos, 254, 259 sq. 

Pores, 75. 

Practical life, 42. 

Problem, 222. 

Prodikos, II 8. 

Proldos on etd&v 0fXot, 91*. 

Protagoras, no ty?., 238 sqg.\ and 
DemokriUxs, Ii2 a , 194, 197 ^.; and 
Zeno, 82, 114 sq.\ and Sokrates, 


Purgation, 41. 
Purifications, 31, 41, 71, 
Pyramid, 6 H . 

Pyrilampcs, 206, 207, 20$, 210. 
Pythagoras, 36-56, 
Pythagoreans, later, 87-93, 289 j$y., 

315, 328, 331 ; d$&v <t>i\ot, 280. 
Pythagorists, 88. 



Quadratrix, 118. 

Rarefaction and condensation, 25. 

Ratio, 47. 

Reality, problem of, n. 

Rebirth, 43, 71. 

Reminiscence, 43, 157 sqq. 

Renaissance, 217. 

Respiration, cosmic, 25, 44, 67, 73- 

Rhetoric, 119. 

Rhind, papyrus, 7. 

Rings, planetary, 24, 56. 

Roman Law, 303 sq. 

Roots, 72. 

Rotation, diurnal, 74. 

Sabazios, 31. 

Sardeis, fall of (546 B.C.), 19. 

Saving appearances, II. 

Scales, 46. 

Science and philosophy, 11-13. 

Seeds, 77. 

Sensation, 75 196 sq., 238 sq. 

Sensible and intelligible, 89^., 159, 164. 

Seven Wise Men, 18. 

Simmias, i$i sq. 

Sokrates, 126-192, 64, 90, 124, 186 

sqq,, 236. 

Sokrates the younger, 238. 
Solar anomaly, 346. 
Solids, regular, 89, 323. 
Sophists, 105-122, 170, 273. 
Soul, 25, 29, 31, 42, 59, 62, 63, 92, 

153, 160, 161, 166, 177, 333 W- 
Space, 51, 67, 343 sq. 
Speusippos, 205, 223, 298, 324, 350. 
Sphere, 55. 

Spheres, * harmony ' of, 56. 
Stars, 24, 36. 

Stereometry, v. Geometry, Solid. 
Stewart, Prof. J. A., 1 68. 
^ulva-sutra, 9. 
Sun, 24, 36, 75, So, 227, 
Surds, 83, 85, 238, 321. 
Survival of the fittest, 24. 

Tarantism, 41. 

Taras, 87. 
Taureas, 190. 

Taylor, A. E., 5O 1 , 85*, 1842, 191. 
Temperament, 50. 
Temperance, 50. 
Temperature, 50. 
Terms, 48. 
Tetraktys, 52. 
Thales, 18-21. 
Theaitetos, 89, 225, 237 sq. 
Thebes, 300. 
Theodores, 211, 238. 
Theophrastos, 347. 
Theoretic life, 42. 
Thourioi, 71, 86, 106, in. 
Tbrasymachos, 121 sq. 
Thucydides (i. 6), 35. 
Time, 342 sq. 
Tranquillity, 199. 
Transmigration, 43. 

Triangles (3:4: 5)> 20, 40, 54 ; (iso- 
sceles right-angled), 54, 83, 89, 156. 

Unit, 83, 321 sqq. 

Unlimited, 44, 51, 83. 

Up and down, 23, 74 sq., 96. 

Voice of Sokrates, 183 sq. 
Void, 95. 
Vortex, 99. 

Weight, 96, 97, 100. 

Worlds, innumerable, 23, 25, 99. 

Xanthippe, 129. 
Xenokiates, 340, 344 sq. 
Xenophanes, 33-36. 

Xenophon and Sokrates, 126 sq., 147 

sqq.y 185 ; Memorabilia (i. 2, 12 sqq.) 

187, (i. 2, 48) 151, (i. 6, 14) 148, (Hi. 

1 6, i) 207, (iii. 7) 210, (iii. 11, 17) 

148, 152, (iv. 5, 12) 228, (iv. 6, 13) 

149, (iv. 7, 3-5) H8 ; Apology (29) 

Zagreus, 31. 

Zeno, 82-85, 82 sq., 89, 114 sq., 134 
sg., 156. 


, 34- 
, v. Good. 

, 221. 
, 21 ; . Air. 
21, 78. 
ts, 238 ^. 
, 174. 

of Protagoras, 113. 

is, 62, 

peiv (u7r5^ecrtf), 163, 229. 
dvdXvcrts, . Analysis. 

i<Wi v- Reminiscence. 

dvdp&triva, (fipovelV) 29 
$, 99. 

a, Il6, 275, 276. 
, 92. 

os dpx^j 2 3- 
r, 22, 39, 44, 51, 90. 
c K<xi Xi)cras, 222. 
^, 75. 

rts, 285 j^/,, 288, 289, 330. 
, w. Goodness. 

$, 48. 

) Number, 
vta, 45, 49, 50, 56, 62, 92, 177. 

dpTrjpta, 77 l . 

, 333 -W- 
, 76, 8l, 112. 

t dpt^/xof, 3 14. 
, 320 ; rpfr?; a%, 225. 

, 31. 

, v. Becoming. 

^opd, 70, 76, 161 J^. 

, 28. 
5a//xw^, 199. 
5e(/ceXa, 196. 
fafrrcpos TrXoOs, 162, 292. 
5iei#e<rts, 325 ^. 
5ta/pe<7i?, 220. 
, 289. 

s, 253. 
5ta'?7, 27. Justice. 
5^7, z/. Rotation, Vortex. 
SiTrXdcnos 1 X^yor, 53. 
5icrcro X6yoi, 231. 
5o|a (dist. arurr///^), 172, 173 -W- J (^ n 

sense of judgement), 248 jyf/., 287 

*/</., 289. 

%, 49, 50, 51, 52, 53, 88 1 , 90, 9I 1 * 

119, 154, .r^/., 196. 
cl'5wXa, 196. 
'eZffwjr^Xot, 91, 279^., 3I3 1 . 

t 132. 

i, 2$4 l > 3^5) S 1 ^- 

icsjy, r6 268. 

325 ^. 

a. TT}S oiV^as, 232, 
Mrpiros X6yos, 53. 
$, 99. 



0S, 47. 

L, 52. 

ta, 199, 325. 
) 199 sq. 

s \6yos, 53* 

6dov, r6, 29, 32. 

0eos, 0eo, 28 ; z>. God, gods. 

0<(0-is, 106. 

deupelv, 42. 

i'5<?ct, 88 1 , 98, 154 sgq. 
u, 348. 
, 50. 
icrrop/77, 38, 58. 

s, 320. 
, 31, 7*- 

KddcLpffLS, 41. 
Ka.Oa.pTal, 32. 

wara/3(XXco, II3 2 , 198, 231. 

^, 1 66. 
/cotvci, 247. 
Kowuvla,, (of forms with sensibles) 165 ; 

(of forms with forms), 225 j^. ; 

282 j^j/. 
Ac6cr/xos, 23 1 . 

/fpao-w, 48, $0 74, 177, 329- 
/cpar^p, 49. 
Aru/cXo<o/>(a, 33 5 1 - 

s, 174. 

X670S (speech, language), 287 ^- ; 
(* Word'), 58 ; events ^ X670", 146, 
162, 282, 315, 3i7 a > 327; ^70" 
h86v<M, 10, 228 ; /xerA X670U, 174, 
250 $g., 252 sg. ; ratio, 47, S3 74, 


See also 54<r<rol X67<u, &vri\oyla. 

, 31, 

, rd, 314, 315 J^. 
, #. Mean. 

ra, 315 sqq. 
, 24. 

&yav t 30. 
L, 26, 78. 

, 83, 321. 

a, 83. 

Z/cos, 72 J^. 

7 45 1 - 

Qeots, l8o 2 . 

1 06, 302 J^. 
6/UOS (dist. 0i5crts), 105 J^. 
6//y (dist. re]}), 197- 
ous, 79- 


^o^a (dist. p^a), 287. 
6'pos, 48, 54, 328. 

ovpavbs, 23 1 . 

ouoTia, ^. Being and becoming. 


42, 2O0 1 . 
Trapouo'tct, 165* 
s, v. Limit. 
oTT-f), 244 1 . 
c6p?7<ns, 80. 
Tjrai, 8. 
t, 46. 

r, 245, 287, 289. 
7r<5pot, 74, 75, 196. 

7Tp6/3X77^a, 222, 226. 
TTpOofy-UCtj 3OI. 

7rp6ra<rts, Trporelj/w, 222. 
s, 6 8 . 

s, ol y 245. 
^i5crts, 322'. 

s, 2OO 1 . 
, II. 

S, I08 2 , 22? 1 . 


<TTdcnwrcu rou 6'Xou, 246. 


ict, 61, 72, 88, 97, 251, 252, 341. 

J, 72, 86. 

vra, 163, 261. 

, 45- 

, 39. 
<rvvex.&, 83, 90- 
<ru?5*>, r6, I 1 , 222. 

cr^etv rA, </>at^6ftc^a, II. 
icrco^a cr7}/xa, 31, I3 1 " 

, 1 1 8. 


rptros Av^pwTros, 254, 259 f. 
rp6iroi, 49- 

y, 45 1 ' 

, 162 s$. ,222, 

z;. Hypothesis. 
(fravTCiffta, 289. 

<f)O.iv6fJLVCL ) T<, II. 

(f>dcn.s, 289. 

<pB6yyoi eery-tores, KWOVJACVOI, 46. 

^tXfa, 72 j#. 

iXo(ro0^a, v. Philosophy. 
iJrt$, 27, 74 1 , 105. 


, 78. 

VOTtS, 62. 

t6pa, 54. 

s, 165, 167, 262, 314* 

<, 55 1 * 90. 

clx^^Xt^ov, 243, 248. 


Recent Philosophical Works 

Prolegomena to an Idealist Theory of Know- 
ledge. By Norman Kemp Smith, D.Phil, LLD, Pro- 
fessor of Logic and Metaphysics in the University of Edin- 
burgh. 8vo. los. 6d. net. 

A Commentary to Kant's Critique of Pure 

Reason, By Professor Norman Kemp Smith, D.Phil., 
LL.D. New Edition, revised, with an Appendix on Kant's 
" Opus Postumum." 8vo, 255. net. 

Mens Creatrix. An Essay. By the Right Rev. 
William Temple, D.D., Lord Bishop of Manchester. Third 
Impression. Svo. los. net. 

Christus Veritas. An Essay. By the Right Rev. 
William Temple, D.D., Lord Bishop of Manchester. Svo. 

The Development of Berkeley's Philosophy. 

By G. A. Johnston, M.A., D.Phil., sometime Lecturer in 
Moral Philosophy in the Universities of St. Andrews and 
Glasgow. Crown Svo. 8s. 6d. net. 

The Philosophy of Hegel. A Systematic Exposi- 
tion. By W. T. Stace, B.A., author of " A Critical History 
of Greek Philosophy." Svo. iSs. net. 

The Philosophic Basis of Moral Obligation. 

A Study in Ethics. By J. E. Turner. Svo. I2s. 6d. net. 

The Beautiful. By Henry Rutgers Marshall, 
L.H.D., D.S., author of " Pain, Pleasure, and Aesthetics," 
etc. Svo. 153. net. 

Freedom of the Mind in History. By Henry 
Osborn Taylor, Lilt.D., author of " The Mediaeval Mind/' 
etc, Crown Svo. 73. 6d. net. 


Recent Philosophical Works 

The Nature of " Intelligence " and the Prin- 
ciples of Cognition. By C. Spearman, Ph.D., Grote 
Professor of Mind and Logic in the University of London. 
8vo. 155. net. 

The Scientific Approach to Philosophy. Selected 
Essays and Reviews. By H. Wildon Carr, D.Litt., Pro- 
fessor of Philosophy in the University of London. 8vo. 

A Theory of Monads. Outlines of the Philosophy 
of the Principle of Relativity. By H. Wildon Carr, D.Litt. 
8vo. 153. net. 

The General Principle of Relativity in its 

Philosophical and Historical Aspect. By H. Wilclon 
Carr, D.Litt. Second Edition. With a new Chapter on 
" Einstein's Theory." Crown 8vo. 73. 6d. net. 

Knowledge and Truth. An Epistemological Essay. 
By Louis Arnaud Reid, M.A., Lecturer on Philosophy in 
University College of Wales, Aberystwyth. 8vo. tos. 6d. 

The Contact between Minds. A Metaphysical 
Hypothesis. By C. Delisle Burns, Author of " Political 
Ideals," etc. Svo. 73. 6d. net. 

The Theory of Mind as Pure Act. By Giovanni 

Gentile, Professor in the University of Rome. Translated 
from the Third Edition, with an Introduction, by II. Wildon 
Carr, D.Litt. 8vo. 153. net. 

A Faith that Enquires. The Gifford Lectures 
delivered in the University of Glasgow in the years 1920-21. 
By Sir Henry Jones, M.A., LL.D.,' D.Litt. 8vo. x8s, net. 

The New Idealism* By May Sinclair. Svo. 143. 



Other Philosophical Works 


Creative Evolution. Translated by Arthur Mitchell, 
Ph.D. 8vo. I2s. 6d. net. 

Laughter : An Essay on the Meaning of the Comic. 

Authorised Translation from the Sixth Edition by Cloudesley 
Brereton, L. es L. (Paris), M. A. (Cantab.) ; and Fred Rothwell, 
B.A. (London). Extra Crown 8vo. 6s. net. 

Mind -Energy : Lectures and Essays. Translated 

by Prof. H. Wildon Carr. 8vo. los. net. 

Henri Bergson. An account of his Life and 
Philosophy. By Algot Ruhe and Nancy Margaret Paul. 
With Portrait. Extra Crown 8vo. 6s. 6d. net. 

The Philosophy of Change. A Study of the 
Fundamental Principle of the Philosophy of Bergson. By 
Prof. H. Wildon Carr. 8vo. 75. 6d. net. 

A Critical Exposition of Bergson's Philosophy. 

By J. M'Kdlar Stewart, B.A., D.Phil. Extra Crown 8vo. 
6s. 6d. net, 

Bergson and Future Philosophy. An Essay on 
the Scope of Intelligence. By George Rostrevor. Crown 
Svo. 73. 6d. net. 


Philosophy of the Practical. Economic and 
Ethic. Translated by Douglas Ainslie, B.A. 8vo. 173. net. 

Aesthetic as Science of Expression and General 

Linguistic. Translated by Douglas Ainslie, B.A. Second 
Edition. 8vo. sis. net. 
Logic as the Science of the Pure Concept. 

Translated by Douglas Ainslie, B.A. 8vo. 173. net. 

What is Living and what is Dead of the Philo- 
sophy of Hegel. Translated by Douglas Ainslie, B.A. 
8vo. 123. 6cl. net. 

The Philosophy of Benedetto Croce : The Prob- 
lem of Art and History. By Prof. H. Wildon Carr. 8vo. 
73. 6d. net. 


Other Philosophical Works 

The Value and Destiny of the Individual. Gifford 

Lectures for 1912. Second Edition. 8vo. 123. 6d. net. 

Three Lectures on Aesthetic* Extra Crown 8vo. 

43. 6d. net. 

The Philosophical Theory of the State. Fourth 

Edition. 8vo. 153. net. 

The Essentials of Logic. Crown 8vo. 43. net. 
Some Suggestions in Ethics. Crown 8vo. 6s. net. 

Implication and Linear Inference, Crown 8vo. 
73. 6d. net. 

What Religion Is. Crown 8vo. 33. 6d. net. 

The Meeting of Extremes in Contemporary 

Philosophy. New and Cheaper Impression. Crown 8vo. 
6s. net. 

Three Chapters on the Nature of Mind. Extra 

Crown 8vo. 6s. net. 


A History of Modern Philosophy : A Sketch of 
the History of Philosophy from the Close of the Renaissance 
to our own Day. Translated by B. E. Meyer. Two vols, 
8vo. i8s. net each. 

The Philosophy of Religion. Translated by B. E. 

Meyer. 8vo. 153, net. 

A Brief History of Modern Philosophy. Trans- 
lated by C. F Sanders. Crown 8vo. 93, net. 

Outlines of Psychology. Translated by M. E. 

Lowndes, Crown 8vo. 73. net. 

Modern Philosophers : Lectures ; and Lectures 
on Bergson, Translated by Alfred C. Mason, M,A, Crown 
8vo. 73. 6d. net.